Conseil Scientifique et Technique du Service de Physique

Conseil Scientifique et Technique duService de Physique Nucléaire

June 3 and 4, 2002CEA/Saclay, DSM/DAPNIA/SPhN,Orme des Merisiers, Building 703

Contents

Contents 1

Agenda 3

List of members 5

Status report on AGATA 7

Photofission, target tests at SAPHIR 9

Status report on ALICE and PHENIX 77

Status report from Exotic Nuclei group 85

Megapie 91

The TRADE project 101

Status report on COMPASS 119


June 3 and 4, 2002CEA/Saclay, DSM/DAPNIA/SPhN,Orme des Merisiers, Building 703

Agenda

Public Session (building 703, room 135)Monday, June 313:30 – 14:25 Introduction Nicolas Alamanos (45' + 10')14:25 – 14:45 Status report on AGATA Wolfram Korten (15' + 5')14:45 – 15:20 Photofission (letter of intent) Henri Safa (25' + 10')15:20 – 15:50 Status report on ALICE and PHENIX Alberto Baldisseri (20' + 10')15:50 – 16:10 Coffee break16:10 – 16:40 Status report from Exotic Nuclei group Laurent Nalpas (20' + 10')16:40 – 17:15 MEGAPIE (new proposal) Frédéric Marie (25' + 10')

Public Session (building 703, room 135)Tuesday, June 3 9:00 – 9:35 The TRADE project (letter of intent) Samuel Andriamonje (25' + 10') 9:35 – 10:05 Status report on COMPASS Yann Bedfer (20' + 10')10:05 – 10:30 Coffee break

Closed Session (building 703, room 125)Tuesday, June 410:30 – 12:30 Closed Session12:30 – 14:00 Lunch break14:00 – 16:30 Closed Session


Members

Membres de droit:

Nicolas Alamanos Michel SpiroChef du SPhN Chef du DAPNIACEA/Saclay CEA/SaclayDSM/DAPNIA/SPhN DSM/DAPNIAF-91191 Gif-sur-Yvette F-91191 Gif-sur-YvetteFrance France

Membres élus:

Michel Garçon (chairman) Frank Gunsing (secretary) Wolfram KortenCEA/Saclay CEA/Saclay CEA/SaclayDSM/DAPNIA/SPhN DSM/DAPNIA/SPhN DSM/DAPNIA/SPhNF-91191 Gif-sur-Yvette F-91191 Gif-sur-Yvette F-91191 Gif-sur-YvetteFrance France France

Fabienne Kunne Danas RidikasCEA/Saclay CEA/SaclayDSM/DAPNIA/SPhN DSM/DAPNIA/SPhNF-91191 Gif-sur-Yvette F-91191 Gif-sur-YvetteFrance France

Membres nommés:

Wolfgang Bauer Alex Brown Piet van DuppenNSCL NSCL Inst. Kern- en StralingsfysicaMichigan State University Michigan State University Department Natuurkunde enEast Lansing, MI 48824-1321 East Lansing, MI 48824-1321 SterrenkundeUSA USA University of Leuven

Celestijnenlaan 200 DB - 3001 LeuvenBelgium

Dietrich von Harrach Marek Lewitowicz Yuri OganessianInstitut für Kernphysik GANIL Flerov Lab. of Nuclear ReactionsJoh. Gutenberg Universität BP 55027 JINRJ. J. Becher Weg 45 F-14076 Caen Cédex 141980 Dubna, Moscow regionD-55099 Mainz France RussiaGermany

Invités permanents:

Françoise Auger Jean-Paul Blaizot Paul Bonche Adj. au Chef du SPhN Chef du SPhT CEA/SaclayCEA/Saclay CEA/Saclay DSM/SPhTDSM/DAPNIA/SPhN DSM / SPhT F-91191 Gif-sur-YvetteF-91191 Gif-sur-Yvette F-91191 Gif-sur-Yvette FranceFrance France

Daniel Guerreau Yves TerrienDirecteur Adjoint Assistant du DirecteurIN2P3 CEA/Saclay3, Rue Michel-Ange DSM/DIRF-75781 Paris Cédex 16 F-91191 Gif-sur-YvetteFrance France

Alban Mosnier Pierre-Olivier Lagage François DamoyChef du SACM Chef du SAP Chef du SDACEA/Saclay CEA/Saclay CEA/SaclayDSM/DAPNIA/SACM DSM/DAPNIA/SACM DSM/DAPNIA/SDAF-91191 Gif-sur-Yvette F-91191 Gif-sur-Yvette F-91191 Gif-sur-YvetteFrance France France

Philippe Rebourgeard Pierre-Yves Chaffard Pascal DebuChef du SEDI Chef du SIS Chef du SPPCEA/Saclay CEA/Saclay CEA/SaclayDSM/DAPNIA/SEDI DSM/DAPNIA/SIS DSM/DAPNIA/SPPF-91191 Gif-sur-Yvette F-91191 Gif-sur-Yvette F-91191 Gif-sur-YvetteFrance France France

Conseil Scientifique et Technique du SPhN

RESEARCH PROPOSAL

Title: The Advanced Gamma Tracking Array : AGATA

Experiment carried out at: GANIL, GSI, Legnaro, Jyväskylä and others

Spokes person(s): not applicable

Contact person at SPhN: W. Korten

Experimental team at SPhN: Y. Le Coz, Ch. Theisen

List of DAPNIA divisions and number of people involved: SEDI

People to be decided after further discussions

List of the laboratories and/or universities in the collaboration and number of people involved:

~35 laboratories and > 200 physicists and engineers from France, Germany, Italy, UK and

others are currently involved in the collaboration.

SCHEDULE

Possible starting date of the project and preparation time [months]: 2002 – 2004 (R&D)

Total beam time requested: not applicable

Expected data analysis duration [months]: not applicable

REQUESTED BUDGET

Total investment costs for the collaboration: approx. 4 M€ for the 1st phase (2002-2004)

Share of the total investment costs for SPhN: 326 k€ for the 1st phase (2002-2004)

Investment/year for SPhN: 111 k€ in 2002, 225 k€ within 2003/4

Total travel budget for SPhN:

Travel budget/year for SPhN:

If already evaluated by another Scientific Committee:

If approved Allocated beam time: Possible starting date:

If Conditionally Approved, Differed or Rejected please provide detailed information:


LETTER OF INTENT

Title: SPIRAL II – Photofission

Target Test at SAPHIR (CEA Saclay)

SPIRAL II Project at GANIL (Caen)

Experiment carried out at: Saclay (target test) & GANIL (location of project)

Spokes person(s):

Contact person at SPhN: Henri Safa

Experimental team at SPhN:

A. Letourneau, F. Marie, D. Ridikas, H. Safa (target test)

List of DAPNIA divisions and number of people involved:

For target test : SPhN, SIS, SACM, approx. 12 people


For target test : DAPNIA, DIMRI (CEA Saclay), IPN (Orsay), GANIL (Caen), approx. 30 people

SCHEDULE

Estimated total duration of the proposed experiment:

Target Test : 1 year

SPIRAL II project : 3 years

Possible starting date of the experiment:

Target Test : January 2003

SPIRAL II project : January 2003

Expected duration of the data analysis: 3 months

ESTIMATED BUDGET

Total investment costs for the collaboration:

Target Test : 250 k€

SPIRAL II project : 20 M€

Share of the total investment cost for SPhN: ?

Total Travel Budget for SPhN: ?

Introduction :

The SPIRAL II project has been initiated by the GANIL (Grand Accélérateur National d'IonsLourds, in Caen) and is aiming at producing and accelerating heavy neutron rich exotic ion beams. In thisproject, neutron rich nuclei from Z=28 to Z=60 could be formed by the fragmentation of a heavy target(Uranium). The study of nuclear matter in this domain of the chart nuclei is still unexplored. A lot of newphysics maybe expected to come out from these new type of ions (see the physics case in [1]). Apart fromidentifying all these new elements, the study of the nucleus quite far from the stable valley, the r-processof nucleosynthesis in astrophysics or the possible superheavy element production are some examples ofthe exciting new nuclear physics that could be reached with SPIRAL II.

Fission fragment production through photofission

One way to produce the Radioactive Ion Beams (RIB) is photofission. Photon having an energy ofroughly 15 MeV are able to provoke fission in a heavy element (like for example uranium 238). Thisphotofission process may exhibit large enough cross section (a few hundreds of mb) to be attractive ascompared to standard neutron induced fission. The photon flux can be created using a 45 MeV electronaccelerator. This is a very important issue as electron machines are quite compact and much cheaper thana deuteron accelerator having the same energy. Electrons are stopped in a heavy Z target to be convertedto photons by bremsstrahlung. The two process (conversion and photofission) could be advantageouslymerged using a single uranium target. In the following description, we will sum up the differentcharacteristics of this project. This information is extracted from the reference [1] : The SPIRAL IIProject, Preliminary Design Study.

Study of a target for photofission

The aim of this work is to design and fabricate a target for the RIB production, optimized for thephotofission process. The design is done using nuclear codes like MCNPX with photonuclear capability.The target study include diffusion/effusion analysis and a complete thermal calculation. In order torelease the exotic atoms created, the target will have probably to be heated to temperatures as high as2000°C. At these temperatures, the atoms can diffuse in the bulk up to the surface, evaporate and theneffuse to the ion source. All these physics phenomena have to be thoroughly addressed to properlyoptimize the target design. The aim is not only to get the higher production rate, but also to try to releasemost efficiently the highly exotic atoms.

The experimental part of the work is to build the SPIRAL II target. That assumes a full masteringof the fabrication process and the mechanics. Then an irradiation test could be performed using a realelectron beam of the SAPHIR facility located at CEA Saclay. This electron machine is able to produce a30 MeV electron beam to the target (See attached Document N° 2 for the detailed SAPHIR experiment).This test will not only demonstrate the target feasibility but would also enable to study its behavior underirradiation. The fission rate could be experimentally measured and compared to theoretical prediction.Moreover, a cryogenic cold finger could even measure some exotic elements production (like rare gases).This SAPHIR test would fully validate the SPIRAL II photofission target. The whole experiment isdescribed in the attached document : The Target Test at SAPHIR : Proposed Experimental Set-up.

Description of the Photofission project

Photofission

It has recently appeared that photofission could be an alternative to n-induced fission. This working grouphas therefore initiated a study of photofission induced by Bremsstrahlung generated by electrons. Therespective merits and technological challenges of these two methods will be evaluated in order to madethe best choice for SPIRAL-II.

With an electron driver, electron interaction with matter will radiate Bremsstrahlung photons inside thetarget. Fission will then be induced by those photons exciting the Giant Dipolar Resonance (GDR) of thenucleus at the right energy. This well-known process is called photofission.The GDR cross section for 238U is shown in the figure below on the left. A maximum fission probabilityof 160 mb is obtained for photons having energy around 15 MeV. At that energy, the photoelectric andthe Compton and Rayleigh scattering cross sections are starting to fall off rapidly so the maincontributions to gamma absorption are e+e- pair production and the photonuclear reactions (γ,f), (γ,n) and(γ,2n). Although the absolute fission cross section is rather small (compared to normal fission withneutrons), its contribution is not negligible as even a pair production reaction may in a thick targeteventually lead to a fission through the resulting photon produced. In the same manner the neutronsproduced by (γ,n) and (γ,2n) reactions as well as the (γ,f) itself can also induce fission, this time by theregular (n,f) high cross section (0.5 barn for fast neutrons). Therefore, in a thick target, photofission maybe a rather interesting way of creating radioactive fission fragments.

0

50

100

150

200

0 5 10 15 20 25 30

Photon Energy (MeV)

σσσσ (

mb

arn

)

U238 Exp

U238Theoretical fit

Photofission cross-section for 238U above (left and right)and rate of fission per electron according to incident energy of the electrons (right) [4].

Unfortunately, no efficient monochromatic sources of 15 MeV photons are available. The most commonway for producing high gamma fluxes is the Bremsstrahlung radiated by passage of electrons throughmatter. This process has a cross section rising linearly with energy. It will dominate the ionization processabove a critical energy (around 20 MeV). But the resulting Bremsstrahlung spectrum is widely spread inenergy from zero up to the full initial electron energy. Although each single electron may ultimatelyproduce as high as 20 photons, only a small fraction of it (0.5 to 0.7 gamma per e-) are "useful" photonslying in the GDR range (15 ±5MeV).A simple calculation including the main electron interactions (Bremsstrahlung and ionization) and themain nuclear reactions of interest (pair production and fission cross section) in a thick depleted uraniumtarget can give the expected number of fission per incident electron (third figure above). We can also planto use a tungsten converter in front of the 238U target. It appears that when using a converter in theelectron driver option, less than 30% of the beam power is lost inside the converter (in contrast to the ddriver option). In the direct method one will produce about 25% more fission per electron (and probablymore when taking in account neutron induced fission). Fission production is almost linear above athreshold energy of 10 MeV. High production is obtained above 40 MeV.As the SPIRAL II project is aiming at 1013 fissions/s, the required beam intensity should be 500 µA foran electron energy of 45 MeV directly on the target.

The Electron Driver Accelerator for SPIRAL II

The accelerator layout is shown below and is quite similar to the MACSE project [2] The injectorcomprises a 100 keV gun and a short cryomodule containing a single superconducting cavity. At theinjector exit, the electron beam is already well relativistic. Then the beam runs in a long cryomodulethrough four SCRF cavities bringing the electrons to the final energy of approximately 45 MeV. Abending magnet followed by a transport beam line will get the beam at the right place and shape onto thetarget.

InjectorInjector

collimatorcollimator

ββββββββ=1=1CryomoduleCryomodule

beambeam dump dump

analysisanalysis

targettarget

beambeamdumpdump

transporttransportlineline

deviationdeviation

100keV100keV 5MeV5MeV 45MeV45MeV

4 SCRF4 SCRF cavities cavitiese-e-gungun capture SCcapture SC cavity cavity

12 m

Layout of the electron driver.

Production Target

The production target is a critical point in several aspects :- It has to sustain the beam heating. In this concern, a converter can be used to dissipated part of the

heat, but with some cost in fission production.

- It raises a great problem of radioprotection, due to the beam stopping and the activation of the target(while the electron driver only causes radiological gamma rays).

- It has to be optimized for the fastest and most efficient effusivity of the fission fragments. Because ofthe adaptability of the electron beam shaping, different targets geometry can be considered, forexample an annular target to facilitate the emission of the atoms.

These different problems will be at the heart of our theoretical and experimental study of the target.

Comparison between photofission and neutron induced fission fragment production

A full description of the LINAG project using neutron-induced fission with a 40MeV linear deuterondriver can be found in [3].In order to compare rapid neutron induced fission and photofission, measurements of Kr and Xe isotopicdistributions produced by photofission and diffused out of a thick UCx target has been performed usingthe same PARRNe1 device in the same conditions that with deuterons beams and with the same target.The measurements have been done with a 4 mm W converter in different position (8 mm from the target,4 mm from the target) and one measurement without W converter. Comparison with the 80 MeVdeuteron induced fission measurements are presented below.

The results obtained are well understood taking into account the percentage of photons between 11 and 17MeV emitted in the cone subtended by the target that is the solid angle.Finally, the above figure summarizes some comparison of production yields normalized to 1013 fissionsfor 100 MeV deuterons (LAHET calculations) and 50 MeV electrons based of the K-H Schmidtestimations. It should be noted that, while intermediate mass nuclei are favoured with high energydeuteron, this discrepancy disappeared with 40MeV deuteron. This makes both modes of production verysimilar as far as production rates are concerned.

Post acceleration

Several configurations have been studied for the acceleration of the fission fragment, including the use ofthe existing GANIL facilities or completely new ones. We sum them up here :

Post acceleration by CIME of the radioactive beams generated through fission of uranium targetsis limited to energies below 10 MeV/n. According to the ion species and to the possibility to go to highercharge states at the expense of the intensity, there might exist an energy gap between the CIME maximumenergy and the SSC2 minimum one.

Energies above 21 MeV/n can be reached using the three GANIL cyclotrons as a post-accelerator.However, we want to stress the point that tuning and controlling several cyclotrons in a row for verysmall intensity beams is not at hand. No strong experience is acquired yet, even on a single cyclotron likeCIME and this procedure seems very un-appropriate.

The postacceleration using CIME and the combination of CIME + CSS2 is therefore limited below 21MeV/n as does not allow to cover the whole programme of physics described in the beginning of thereport.

The injection of the fission fragments through the whole GANIL accelerator (C0+CSS1+CSS2) could bealso considered and would cover the present GANIL energies, but the transmission efficiency has to beevaluated.

The possibility of adding to the SPIRALII project a linear post accelerator up to energies of 100 MeV/u isunder discussion and consideration. This would allow to cover the wide programme of physics.

Safety aspects

Up to now, few works have been carried out to evaluate the safety constraints and their influence on theSPIRAL II layout. The only data available today concern the safety of the electron linear accelerator (a2m thick concrete wall around the accelerator seems sufficient) and the safety around the targets fordeuterons (neutrons flux, fission products, ...).One of the major problems not yet studied concerns the handling and the storage of the irradiated uraniumtargets. Our policy has been to keep open the different options, hoping that the room left free to install theequipments needed to respect the safety constraints would be sufficient. The detailed studies, carried outduring the project design period, will allow to define these points precisely.

SPIRAL II Cost assessment

The cost of SPIRAL II has been devised by taking into account the main following options:• The cost of the buildings has been deduced from the actual cost of the SPIRAL I building (about

2 k€ /m2). This building included an underground floor. For the cyclotron option, an undergroundfloor is needed. For the linac options, the cost should be roughly the same with or withoutunderground floor (the cost of the excavation is compensated by less radioprotection concrete).The basic networks and equipemnts (cranes...) are included.

• The cost of the target-source set and its infrastructures (caves, equipments, target handling system,target storage, ...), estimated at 3.2 M€ , has been extrapolated from the cost of the SPIRAL I one(2 M€ ), taking into account that the target-source set of SPIRAL II will be larger, with moresafety constraints. However, the cost of a target retreatment facility has been considered as anextra-budget (between 2 and 4 M€ according to the level of radioprotection regulation), the basicoption being to store the targets without retreating them.

• The cost of the driver has been devised as follows :! Cyclotron for 80 MeV deuterons: Its price (12 M€ ) has been got from the european

industry. It is valid for a turn-on-key cyclotron. This price could be reduced if somecomponents are taken in charge by GANIL (deuteron source, control system for instance).A participation to the installation could reduce the price also.

! Linear accelerator for 45 MeV electrons: The design and the construction will be carriedon by French laboratories. The salaries of people working in these laboratories are nottaken into account in the cost. Consequently, its price (6.1 M€ ) includes mainly thesupplying of the components. Morever, the re-use of a part of MACSE (prototype of asuperconducting linear accelerator built in Saclay) could reduce the cost (between 1 to 2M€ could be saved).

! Linear accelerator for 40 MeV deuterons: As for the electron linac, the design and theconstruction will be carried on by European laboratories, that means the salaries of peopleworking in these laboratories are not taken into account in the cost and the price includesmostly the supplying of the components (18 M€ ). Moreover, the RFQ, pre-acceleratingthe beam going out from the source, is still in a conceptual design state. So, its price is notyet precisely estimated.

• The cost of the beam lines depends on their performances:! Beam lines from the driver to the target and from the target to CIME. These beam lines are as

sophisticated as the SPIRAL I ones. So; the same price as for SPIRAL I will be retained forthese lines (75 k€ /m);

! The 1/1000 separator has been estimated at 300 k€ ;! The price of the charge breeder is based on the market price of ECR sources (400 k€ )! The beam line from the second separator output to the LIRAT line junction. This line is just a

transport line. Its price is lower (40 k€ /m);! The beam line from CIME to CSS2 is partly a sophisticated line and partly just a transport

line. Its price reflects this fact (50 k€ /m). The price of the corridor-like building to cover theline with radiologic protection must be added (1 k€ /m2);

• The cost of the control-system is relatively low (300 k€ ) because the drivers are simple to control(constant energy, unique particule).

• The cost of the radioprotection is difficult to estimate. Indeed, up to now, there has been ere noreal study of the constraints induced by the regulation, in particular concerning the use of uraniumin the target. The price which is proposed (1.5 M€ ) is just an extrapolation of the cost of theradioprotection for SPIRAL I (1 M€ ), without taking into account further constraints like aconfining vessel around the target.

• Few further expenses have been taken into account like travels (0,15 M€ ), site roads, car parksand green areas (0,15 M€ ).

• A sum of 10 % of hazards has been added.

Estimation of the costIn the next page, is given the estimated cost in millions of Euros (M € )

SPIRAL II BUDGET

(M€€€€ )40 MeV deuteron

linac45 MeV electron

linac80 MeV deuteron

cyclotron

Building / Infrastructure 5,6 3,2 4,3

Driver 18,6 6,2 12,2

Targets / Sources 1+ 3,2 3,2 3,2

Source N+ 0,4 0,4 0,4

Beam lines 4,6 4,3 4,3

Radioprotection 1,5 1,2 1,5

Control system 0,3 0,3 0,3

Miscellaneous 0,3 0,3 0,3

Hazards (10%) 3,5 1,9 2,6

TOTAL (M€ ) 38,0 21,0 29,1

MACSE re-use -2,3

GRAND TOTAL (M€€€€ ) 38,0 18,7 29,1

target retreatment 3,0CSS2 re-acceleration 4,4

References

[1] The SPIRAL II project, Preliminary Design Study, 28/11/2001.[2] Module Accélérateur de Cavités Supraconductrices à électrons, Internal Report DAPNIA/SEA 92-02,1992.[3] High intensity beams at GANIL and future opportunities : LINAG, G. Auger, W. Mittig, M.H.Moscatello, A.C.C. Villari, GANIL Report R01 02.[4] Y. Oganessian & al., 5th International Conference on Radioactive Nuclear Beams, Divonne, France,April 2002.

PHOTOFISSION TARGET EXPERIMENTAT SAPHIR

CEA Saclay, DSM/DAPNIA/SPhN91191 Gif/Yvette, France

Abstract :

The present proposed experiment is to built a complete target designed toproduce neutron rich radioactive ion beams through the photofission process. Thistarget could be irradiated at the SAPHIR facility located at CEA Saclay. The finalgoal is a full scale demonstration of both the technical feasibility of the target and itsfission product release capability. Some specific radioactive gas elements producedlike Kr or Xe could be measured on line and the results compared with the theoreticalprediction.

1. IntroductionThe SPIRAL II project is aiming at producing heavy neutron rich radioactive ion

beams. The production scheme [1] is based on the fission process induced after excitation ofthe Giant Dipolar Resonance (GDR) of uranium using a 15 MeV gamma. Electronsaccelerated to an energy of roughly 45 MeV are slowed down in a heavy Z target to producethe photon flux through the bremsstrahlung emission. Incidentally, the electron/photonconversion could be done in the very same target as the fission production using a singleuranium target. The aim of this experiment is to design and construct that kind of target.Following, an experimental irradiation of the target can be done to measure both the fissionrate and the production of some radioactive elements.

2. Preliminary ExperimentThe SAPHIR accelerator is an electron machine located at CEA/Saclay. Electrons are

accelerated to an energy from 10 to 35 MeV, in pulses of roughly 2.5 µs and with a repetitionrate varying between 6.25 Hz and 400 Hz. The peak current in a single pulse can be as high as150 mA, which makes an average maximum total beam power of 5 kW.

In order to experimentally evaluate the (γ,f) process in 238U, a first experiment hasbeen set up in 2001 where the beam is first selected in energy (see below) and then sent to apure uranium target (a small thick cylinder of 20 mm length and 7 mm diameter. Two kind ofdetectors has been used. First the "standard" Helium3 neutron detector to study the delayedneutrons. That will give some information on the fission rate. Then a fission detector tomeasure the neutrons produced during the pulse. These prompt neutrons are a signature of thesum of all nuclear interaction mainly (γ,n) (γ,2n) and (γ,f). By combining the two information,the branching ratio between the two process can be given for a thick target and for a givenenergy. The experimental setup may equally give the angular dependence for each process.The experimental measurement can also be compared with the theoretical predictions.

2. 1 - The Transfer Line

Because the electron energy distribution coming out from the SAPHIR accelerator isvery wide, it is not quite well suited to make energy dependence measurements. But thephotofission process is really a threshold process as regard to electron energy with almost noproduction below some 15 MeV. And the production rate rises very steeply in the range 20 to40 MeV. So it is quite important to work with almost monoenergetic electrons. In that sense, aspecific transport line has been built and installed at the SAPHIR facility. A deviating magnetselects the energy by applying a specific magnetic field then the beam is deviated again in theexperimental hall to be injected in the transport line shown below (Figure 1). This transportline will help for all future foreseen photofission experiments.

! " # $ % #

& ' ! ( !

)

# & * + * +

,

- # . # . / ! 0 1

2 $ ! ! ! 3

surrounding the target covering angle ranging from (-135° to +135°). These detectors arebasically used for the delayed neutron measurement.

Another detector can be used to detect the prompt neutron emission. Due to the highflux during the pulse beam, there is no need to have high sensitivity detectors but rather fastrate acquisition due to the short time (the pulse length is only 2.5 µs). A preliminarymeasurement has been tried with fission neutron detectors that gave some informationprovided the discriminator level was set properly due to the high gamma rate dose emitted atthe same time. More work is needed to get a full and proper set of detectors measuring theprompt neutron emission with improved accuracy. A faster acquisition electronics is alsoneeded to separate two neutron signal.

2. 4 - Preliminary Results

The first experimental set-up is shown in Figure 2. The target is inserted in a closedcylindrical evacuated box sitting on a movable table. The neutron detectors are supported on aaluminum structure and positioned around the target. The only angle not covered is the beamdirection (180°).

Figure 2- Experimental set-up showing the neutron detector position around thetarget.

One of the detector response is shown in Figure 3. A few tens of ms after the pulse, theneutrons are counted and the integrated number of neutrons give the total amount of delayedneutrons. These are directly proportional to the actual number of fission in the target. Thenoise level is measured when the beam is on, but deviated from the target. The detectors arecalibrated using a known 252Cf source placed at the target location. As expected, the fissionsignal is completely isotropic and all the seven detectors give the same signal answer towithin 7% (one sigma) for the reference time of the measurement (arbitrarily set to 600 s).

Réponse Détecteur Hélium3

-20

-10

0

10

20

30

40

50

1 51 101 151 201 251 301 351 401 451 501

Canal

No

mb

re d

e C

ou

ps

I147

Figure 3- Signal of delayed neutron on a Helium 3 detector.

Moreover, to check the linearity of the measurements, the beam intensity has beenchanged over a wide range (2 µA to 25 µA) and the measure done at each intensity. Given theuncertainty on the current measurement (the current is measured using a pulsed currentmonitor and a fast scope integrated over 10 µs and averaged over 16 pulses) and theuncertainty on the actual current hitting the target (for each intensity, a new setting of theaccelerator parameters is required and both the beam focusing and the beam line transfer arecorrected), the overall result is quite satisfactory. The fission rate is measured to beproportional to the beam current (see Figure 4)

Linéarité des mesuresE = 29 MeV

y = 10.62x - 223.71

y = 12.52x - 291.81

y = 20.69x - 407.24

y = 15.20x - 357.54

0

500

1 000

1 500

2 000

2 500

3 000

3 500

4 000

4 500

5 000

0 50 100 150 200 250 300

Intensité (mV)

Co

up

s en

600

s

Detecteur 4

Detecteur 3

Detecteur 2

Detecteur 1

Linéaire (Detecteur 4)




Figure 4 – Linearity of detected fission neutrons with beam current.

3. Target Design & Experimental Set-up

3. 1 - Target Geometry

The photofission target should be quite different from the deuteron or proton targetsdeveloped earlier for similar applications [2,3]. Because the heat dissipation and theproduction method (gamma from bremsstrahlung compared to neutron fission) is quitedifferent, a new target geometry has to be envisaged for the photofission experiment.Therefore, in order to demonstrate the performance of the new target, an irradiation test has tobe conducted to assess the calculated foreseen production and thermal management. Thisexperiment is the so-called SAPHIR proposed experiment for the SPIRAL II photofissiontarget.

3. 2 - Experiment Set-up

The proposed experiment is to build a photofission target designed for SPIRAL II thento irradiate it with the 30 MeV electron beam at the SAPHIR facility. A schematic of the set-up is shown in Figure 5. The aim of this experiment is twofold :

− To verify the calculated fission rate evaluated with codes like MCNPX for thatgiven target geometry. In addition to the neutron measurements (both fast anddelayed) using the detectors described above, in-situ measurements of someradioactive species (mainly gases) can be performed. This is done using a specificline where the fission products leaving the target are propagated and then collectedon a finger cooled by liquid nitrogen (Figure 6). This technique has already beenused in previous experiments [4,5] and the same existing apparatus (with someminor adjustments) may be used in the proposed experiment.

− To demonstrate the overall technical capability of the target, demonstrating thefabrication process of the uranium carbide and the remote control including thethermal management and all the related technical issues like pressure, temperature,cooling, beam control and positioning, etc… Given the available power beam atSAPHIR, the thermal issue can also be checked and compared to the thermalcalculation.

Figure 5 – Set-up of the proposed SAPHIR experiment.

Figure 6 – A view of the line collecting the radioactive atoms produced.

4. Target Technology

4. 1 - Uranium Carbide

The technology of the target material is an important issue for the project. In order torelease efficiently the fission products, the target has to be heated to 2000°C. Therefore, aceramic compound of uranium that is able to withstand that temperature without degradationhas to be selected like for example the uranium carbide. But the carbide has manythermodynamic phases depending on temperature and composition. Moreover, themetallurgical form is also of importance as for example the small grain size is better for thediffusion process. As a consequence, a full and proper characterization of the material target(fabrication process, density, porosity, phase composition and proportion, thermal propertieslike thermal conductivity at 2000°C, emissivity, etc…) is thoroughly needed as it maystrongly impact the target design.

4. 2 - Physics and Calculation

The target geometry results from an overall optimization including many differentphysics phenomena. First, the basic nuclear process of fission induced by gamma particles(called photofission). This includes the full neutronic calculation and the related fissionproducts distribution. Although the GDR process has been studied since long time (forexample, the (γ,f) cross section has been very precisely measured [6]), a precise understanding

of the mechanism is yet to be developed. Consequently, the mass and isotope distribution isnot quite well known and the corresponding theoretical prediction are far from reliable. Eventhe distribution of delayed neutron released during that process is not quite known, eventhough the 238U is probably the most studied atom in this regard.

After the production itself, the diffusion and effusion calculation are relying on thetarget technology on one hand (see 4.1) and on the thermal calculation and thermalmanagement (cooling and heater) on the other. Therefore, there must be a strong and deepinterconnected work between the different physics involved (diffusion of atoms inside thetarget, effusion up to the ion source and thermal profile of the target) to finally optimize thefull target geometry. As a result, an example of a target design that could be used for aphotofission target is shown in Figure 7. Note that the target strongly differs from the oneused for deuterons of protons.

Figure 7 – Example of a photofission target geometry.

5. Cost and ScheduleThe overall cost of the target experiment has been estimated to be of the order of

250 k€ including the cost of the SAPHIR beam. The required time for the target design is sixmonths. After which the target should be fabricated and tested off line (including the heatingtests). Then the irradiation experiment could be undertaken within one year.

References

1 "Photofission for the SPIRAL II Project", H. Safa, D. Ridikas & al., Proceedings of International ConferenceISOL'01, Oakridge, Tenessee, USA, March 2001, also DAPNIA/SPhN-01-11 Internal report, March 2001

2 Parrne target, See for example : B. Roussière & al., IPNO DR-02-002, to be published in NIM3 ISOLDE target, U.H. Köster & al., CERN- 2001-075, Radiochem. Acta 89,11-12, p. 749 [2001]4 F. Clapier & al.,"Exotic beams produced by fast neutrons", Phys. Rev. Spec. Topics AB Vol. 1, 013501 [1998]5 F. Ibrahim & al., "Photo-fission for the production of radioactive beam …", to be published in European

Physical Journal [2001]6 A. Veyssière & al., Nuclear Physics, A199, p. 45 [1973]

Introduction :

The SPIRAL II project has been initiated by the GANIL (Grand Accélérateur National d'IonsLourds, in Caen) and is aiming at producing and accelerating heavy neutron rich exotic ion beams. In thisproject, neutron rich nuclei from Z=28 to Z=60 could be formed by the fragmentation of a heavy target(Uranium). The study of nuclear matter in this domain of the chart nuclei is still unexplored. A lot of newphysics maybe expected to come out from these new type of ions (see the physics case in [1]). Apart fromidentifying all these new elements, the study of the nucleus quite far from the stable valley, the r-processof nucleosynthesis in astrophysics or the possible superheavy element production are some examples ofthe exciting new nuclear physics that could be reached with SPIRAL II.

Fission fragment production through photofission

One way to produce the Radioactive Ion Beams (RIB) is photofission. Photon having an energy ofroughly 15 MeV are able to provoke fission in a heavy element (like for example uranium 238). Thisphotofission process may exhibit large enough cross section (a few hundreds of mb) to be attractive ascompared to standard neutron induced fission. The photon flux can be created using a 45 MeV electronaccelerator. This is a very important issue as electron machines are quite compact and much cheaper thana deuteron accelerator having the same energy. Electrons are stopped in a heavy Z target to be convertedto photons by bremsstrahlung. The two process (conversion and photofission) could be advantageouslymerged using a single uranium target. In the following description, we will sum up the differentcharacteristics of this project. This information is extracted from the reference [1] : The SPIRAL IIProject, Preliminary Design Study.

Study of a target for photofission

The aim of this work is to design and fabricate a target for the RIB production, optimized for thephotofission process. The design is done using nuclear codes like MCNPX with photonuclear capability.The target study include diffusion/effusion analysis and a complete thermal calculation. In order torelease the exotic atoms created, the target will have probably to be heated to temperatures as high as2000°C. At these temperatures, the atoms can diffuse in the bulk up to the surface, evaporate and theneffuse to the ion source. All these physics phenomena have to be thoroughly addressed to properlyoptimize the target design. The aim is not only to get the higher production rate, but also to try to releasemost efficiently the highly exotic atoms.

The experimental part of the work is to build the SPIRAL II target. That assumes a full masteringof the fabrication process and the mechanics. Then an irradiation test could be performed using a realelectron beam of the SAPHIR facility located at CEA Saclay. This electron machine is able to produce a30 MeV electron beam to the target (See attached Document N° 2 for the detailed SAPHIR experiment).This test will not only demonstrate the target feasibility but would also enable to study its behavior underirradiation. The fission rate could be experimentally measured and compared to theoretical prediction.Moreover, a cryogenic cold finger could even measure some exotic elements production (like rare gases).This SAPHIR test would fully validate the SPIRAL II photofission target. The whole experiment isdescribed in the attached document : The Target Test at SAPHIR : Proposed Experimental Set-up.

Description of the Photofission project

Photofission

It has recently appeared that photofission could be an alternative to n-induced fission. This working grouphas therefore initiated a study of photofission induced by Bremsstrahlung generated by electrons. Therespective merits and technological challenges of these two methods will be evaluated in order to madethe best choice for SPIRAL-II.

With an electron driver, electron interaction with matter will radiate Bremsstrahlung photons inside thetarget. Fission will then be induced by those photons exciting the Giant Dipolar Resonance (GDR) of thenucleus at the right energy. This well-known process is called photofission.The GDR cross section for 238U is shown in the figure below on the left. A maximum fission probabilityof 160 mb is obtained for photons having energy around 15 MeV. At that energy, the photoelectric andthe Compton and Rayleigh scattering cross sections are starting to fall off rapidly so the maincontributions to gamma absorption are e+e- pair production and the photonuclear reactions (γ,f), (γ,n) and(γ,2n). Although the absolute fission cross section is rather small (compared to normal fission withneutrons), its contribution is not negligible as even a pair production reaction may in a thick targeteventually lead to a fission through the resulting photon produced. In the same manner the neutronsproduced by (γ,n) and (γ,2n) reactions as well as the (γ,f) itself can also induce fission, this time by theregular (n,f) high cross section (0.5 barn for fast neutrons). Therefore, in a thick target, photofission maybe a rather interesting way of creating radioactive fission fragments.

0

50

100

150

200

0 5 10 15 20 25 30

Photon Energy (MeV)

σσσσ (

mb

arn

)

U238 Exp

U238Theoretical fit

Photofission cross-section for 238U above (left and right)and rate of fission per electron according to incident energy of the electrons (right) [4].

Unfortunately, no efficient monochromatic sources of 15 MeV photons are available. The most commonway for producing high gamma fluxes is the Bremsstrahlung radiated by passage of electrons throughmatter. This process has a cross section rising linearly with energy. It will dominate the ionization processabove a critical energy (around 20 MeV). But the resulting Bremsstrahlung spectrum is widely spread inenergy from zero up to the full initial electron energy. Although each single electron may ultimatelyproduce as high as 20 photons, only a small fraction of it (0.5 to 0.7 gamma per e-) are "useful" photonslying in the GDR range (15 ±5MeV).A simple calculation including the main electron interactions (Bremsstrahlung and ionization) and themain nuclear reactions of interest (pair production and fission cross section) in a thick depleted uraniumtarget can give the expected number of fission per incident electron (third figure above). We can also planto use a tungsten converter in front of the 238U target. It appears that when using a converter in theelectron driver option, less than 30% of the beam power is lost inside the converter (in contrast to the ddriver option). In the direct method one will produce about 25% more fission per electron (and probablymore when taking in account neutron induced fission). Fission production is almost linear above athreshold energy of 10 MeV. High production is obtained above 40 MeV.As the SPIRAL II project is aiming at 1013 fissions/s, the required beam intensity should be 500 µA foran electron energy of 45 MeV directly on the target.


The accelerator layout is shown below and is quite similar to the MACSE project [2] The injectorcomprises a 100 keV gun and a short cryomodule containing a single superconducting cavity. At theinjector exit, the electron beam is already well relativistic. Then the beam runs in a long cryomodulethrough four SCRF cavities bringing the electrons to the final energy of approximately 45 MeV. Abending magnet followed by a transport beam line will get the beam at the right place and shape onto thetarget.

InjectorInjector



beambeam dump dump

analysisanalysis

targettarget

beambeamdumpdump


deviationdeviation



12 m


Production Target

The production target is a critical point in several aspects :- It has to sustain the beam heating. In this concern, a converter can be used to dissipated part of the

heat, but with some cost in fission production.

- It raises a great problem of radioprotection, due to the beam stopping and the activation of the target(while the electron driver only causes radiological gamma rays).

- It has to be optimized for the fastest and most efficient effusivity of the fission fragments. Because ofthe adaptability of the electron beam shaping, different targets geometry can be considered, forexample an annular target to facilitate the emission of the atoms.

These different problems will be at the heart of our theoretical and experimental study of the target.

Comparison between photofission and neutron induced fission fragment production

A full description of the LINAG project using neutron-induced fission with a 40MeV linear deuterondriver can be found in [3].In order to compare rapid neutron induced fission and photofission, measurements of Kr and Xe isotopicdistributions produced by photofission and diffused out of a thick UCx target has been performed usingthe same PARRNe1 device in the same conditions that with deuterons beams and with the same target.The measurements have been done with a 4 mm W converter in different position (8 mm from the target,4 mm from the target) and one measurement without W converter. Comparison with the 80 MeVdeuteron induced fission measurements are presented below.

The results obtained are well understood taking into account the percentage of photons between 11 and 17MeV emitted in the cone subtended by the target that is the solid angle.Finally, the above figure summarizes some comparison of production yields normalized to 1013 fissionsfor 100 MeV deuterons (LAHET calculations) and 50 MeV electrons based of the K-H Schmidtestimations. It should be noted that, while intermediate mass nuclei are favoured with high energydeuteron, this discrepancy disappeared with 40MeV deuteron. This makes both modes of production verysimilar as far as production rates are concerned.

Post acceleration

Several configurations have been studied for the acceleration of the fission fragment, including the use ofthe existing GANIL facilities or completely new ones. We sum them up here :

Post acceleration by CIME of the radioactive beams generated through fission of uranium targetsis limited to energies below 10 MeV/n. According to the ion species and to the possibility to go to highercharge states at the expense of the intensity, there might exist an energy gap between the CIME maximumenergy and the SSC2 minimum one.

Energies above 21 MeV/n can be reached using the three GANIL cyclotrons as a post-accelerator.However, we want to stress the point that tuning and controlling several cyclotrons in a row for verysmall intensity beams is not at hand. No strong experience is acquired yet, even on a single cyclotron likeCIME and this procedure seems very un-appropriate.

The postacceleration using CIME and the combination of CIME + CSS2 is therefore limited below 21MeV/n as does not allow to cover the whole programme of physics described in the beginning of thereport.

The injection of the fission fragments through the whole GANIL accelerator (C0+CSS1+CSS2) could bealso considered and would cover the present GANIL energies, but the transmission efficiency has to beevaluated.

The possibility of adding to the SPIRALII project a linear post accelerator up to energies of 100 MeV/u isunder discussion and consideration. This would allow to cover the wide programme of physics.

Safety aspects

Up to now, few works have been carried out to evaluate the safety constraints and their influence on theSPIRAL II layout. The only data available today concern the safety of the electron linear accelerator (a2m thick concrete wall around the accelerator seems sufficient) and the safety around the targets fordeuterons (neutrons flux, fission products, ...).One of the major problems not yet studied concerns the handling and the storage of the irradiated uraniumtargets. Our policy has been to keep open the different options, hoping that the room left free to install theequipments needed to respect the safety constraints would be sufficient. The detailed studies, carried outduring the project design period, will allow to define these points precisely.


The cost of SPIRAL II has been devised by taking into account the main following options:• The cost of the buildings has been deduced from the actual cost of the SPIRAL I building (about

2 k€ /m2). This building included an underground floor. For the cyclotron option, an undergroundfloor is needed. For the linac options, the cost should be roughly the same with or withoutunderground floor (the cost of the excavation is compensated by less radioprotection concrete).The basic networks and equipemnts (cranes...) are included.

• The cost of the target-source set and its infrastructures (caves, equipments, target handling system,target storage, ...), estimated at 3.2 M€ , has been extrapolated from the cost of the SPIRAL I one(2 M€ ), taking into account that the target-source set of SPIRAL II will be larger, with moresafety constraints. However, the cost of a target retreatment facility has been considered as anextra-budget (between 2 and 4 M€ according to the level of radioprotection regulation), the basicoption being to store the targets without retreating them.

• The cost of the driver has been devised as follows :! Cyclotron for 80 MeV deuterons: Its price (12 M€ ) has been got from the european

industry. It is valid for a turn-on-key cyclotron. This price could be reduced if somecomponents are taken in charge by GANIL (deuteron source, control system for instance).A participation to the installation could reduce the price also.

! Linear accelerator for 45 MeV electrons: The design and the construction will be carriedon by French laboratories. The salaries of people working in these laboratories are nottaken into account in the cost. Consequently, its price (6.1 M€ ) includes mainly thesupplying of the components. Morever, the re-use of a part of MACSE (prototype of asuperconducting linear accelerator built in Saclay) could reduce the cost (between 1 to 2M€ could be saved).

! Linear accelerator for 40 MeV deuterons: As for the electron linac, the design and theconstruction will be carried on by European laboratories, that means the salaries of peopleworking in these laboratories are not taken into account in the cost and the price includesmostly the supplying of the components (18 M€ ). Moreover, the RFQ, pre-acceleratingthe beam going out from the source, is still in a conceptual design state. So, its price is notyet precisely estimated.

• The cost of the beam lines depends on their performances:! Beam lines from the driver to the target and from the target to CIME. These beam lines are as

sophisticated as the SPIRAL I ones. So; the same price as for SPIRAL I will be retained forthese lines (75 k€ /m);

! The 1/1000 separator has been estimated at 300 k€ ;! The price of the charge breeder is based on the market price of ECR sources (400 k€ )! The beam line from the second separator output to the LIRAT line junction. This line is just a

transport line. Its price is lower (40 k€ /m);! The beam line from CIME to CSS2 is partly a sophisticated line and partly just a transport

line. Its price reflects this fact (50 k€ /m). The price of the corridor-like building to cover theline with radiologic protection must be added (1 k€ /m2);

• The cost of the control-system is relatively low (300 k€ ) because the drivers are simple to control(constant energy, unique particule).

• The cost of the radioprotection is difficult to estimate. Indeed, up to now, there has been ere noreal study of the constraints induced by the regulation, in particular concerning the use of uraniumin the target. The price which is proposed (1.5 M€ ) is just an extrapolation of the cost of theradioprotection for SPIRAL I (1 M€ ), without taking into account further constraints like aconfining vessel around the target.

• Few further expenses have been taken into account like travels (0,15 M€ ), site roads, car parksand green areas (0,15 M€ ).


Estimation of the costIn the next page, is given the estimated cost in millions of Euros (M € )

SPIRAL II BUDGET

(M€€€€ )40 MeV deuteron

linac45 MeV electron

linac80 MeV deuteron

cyclotron

Building / Infrastructure 5,6 3,2 4,3

Driver 18,6 6,2 12,2

Targets / Sources 1+ 3,2 3,2 3,2

Source N+ 0,4 0,4 0,4

Beam lines 4,6 4,3 4,3

Radioprotection 1,5 1,2 1,5

Control system 0,3 0,3 0,3

Miscellaneous 0,3 0,3 0,3

Hazards (10%) 3,5 1,9 2,6

TOTAL (M€ ) 38,0 21,0 29,1

MACSE re-use -2,3

GRAND TOTAL (M€€€€ ) 38,0 18,7 29,1


References

[1] The SPIRAL II project, Preliminary Design Study, 28/11/2001.[2] Module Accélérateur de Cavités Supraconductrices à électrons, Internal Report DAPNIA/SEA 92-02,1992.[3] High intensity beams at GANIL and future opportunities : LINAG, G. Auger, W. Mittig, M.H.Moscatello, A.C.C. Villari, GANIL Report R01 02.[4] Y. Oganessian & al., 5th International Conference on Radioactive Nuclear Beams, Divonne, France,April 2002.

November 28, 2001

SPIRAL II : Preliminary Design Study

There is currently in the nuclear physics community a strong interest in the use of beamsof accelerated radioactive ions. Although a fast glance at the nuclide chart immediately shows thevast unknown territories on the neutron-rich side of the valley of beta stability, only few projectsare concerned with the neutron-rich nuclides. The SPIRAL-II program aims at studying thetechniques for delivering beams of neutron-rich radioactive nuclides at energies of a few MeV pernucleon. This energy allows to overcome the Coulomb repulsion between the radioactive beamand the target nuclei in most systems and opens up new possibilities for experimental studies ofneutron-rich nuclei and of the synthesis of the heaviest elements. A number of new phenomenaare indeed predicted to occur in nuclei with large neutron excess which will help to improvenuclear models by comparison with data not available to date. Moreover, it can be noted that theastrophysics community is very interested in nuclear data for calculations of nucleosynthesis.

The scientific council of GANIL asked to perform a comparative study on the productionmethods based on gamma induced fission and rapid-neutron induced fission concerning thenature and the intensity of the neutron-rich products. The production rate expected should bearound 1013 fissions per second. The study should include the implantation and the costs of theconcerned accelerators. The scientific committee recommended also to study the possibility to re-inject the radioactive beams of SPIRAL II in the cyclotrons available at GANIL in order to giveaccess to an energy range from 1.7 to 100 MeV/nucleon

For that purpose, some study groups have been formed to evaluate the possibility of such aproject in the different components: physics case, target-ion sources, drivers, postacceleration andgeneral infrastructure. The organization of the project study is given at the end of this report. Youwill find also the name of all the people involved in the study.

The following report presents an overview of the study.

The physics case of SPIRAL2

The SPIRAL2 facility of GANIL is supposed to deliver high-intensity beams of neutron- richfission fragments. The use of these high-intensity beams at the GANIL low-energy ISOL facilityor their acceleration to a few tens of MeV/nucleon opens new possibilities in nuclear structurephysics, nuclear astrophysics, as well as in reaction dynamics studies.A few of these nuclear physics related subjects will be described here.

Towards the neutron drip line

The existence of an atomic nucleus defined by its stability with respect to the strong interaction isprobably its most basic feature. The limit of stability of neutron-rich nuclei, the neutron drip line,has up to now been experimentally reached only up to oxygen (Z=8). SPIRAL2 beams shouldallow to go to higher-Z elements and to test the limits of stability for these elements. Thesestudies can be conducted by accelerating and fragmenting neutron-rich fission products inprojectile fragmentation-type experiments.Even if the neutron drip line is out of reach for elements above Z=14 or so, projectilefragmentation of neutron-rich nuclei is an interesting way to produce very neutron-rich nuclides.As an example, doubly-magic 78Ni may be produced via projectile fragmentation of 82Ge fissionfragments. In a similar way, 110Zr can be produced from 114Ru secondary beams. Rough estimatesshow that the loss in intensity of secondary fission-fragment beams as compared to stable primarybeams may be overcompensated by the higher fragmentation cross sections. In selected cases, thisshould yield counting rates of exotic fragments one or two orders of magnitude higher thanreached by stable-beam projectile fragmentation. These types of studies need rather high beamenergies for the fission fragments of about 50-70 MeV/nucleon.Decay studies with these neutron-rich isotopes will allow for many interesting studies. Inparticular, neutron-neutron correlation studies in beta-delayed two-neutron emission will enableus to determine the time scale of the neutron emission and the size of the emitting source.The comparison between experiment and theory, in particular for the location near or in the''island of inversion'' (N~20), of Gamow Teller (GT) and natural parity states fed in beta decay,has allowed to obtain crucial information on the neutron-neutron and proton-neutron interactionin the involved subshells. In this context it would be of prime interest to explore other massregions near the shell closures (e.g. Z ~40, 50 and N ~50).

Single-particle levels and spin-orbit splitting around 132Sn

The shell-model description of the structure of the atomic nucleus is based on a picture where allthe nucleons are arranged in shells, each capable of containing a maximum number of nucleons.These shells corresponding to magic nucleon numbers are well established at the stability line andfor radioactive nuclei close to stability. However, it is not possible to predict how this shellstructure evolves far from stability. In particular, it has been shown that, for neutron-rich isotopes

around the classical shell closures at N=20 and N=28, the shell gap between two major shells iseither strongly reduced or intruder configurations become very important.SPIRAL2 is the ideal facility to study the N=82 shell closure over a wide variety of nuclei. Asshown in figure 1, only a few single-particle levels are known in the 132Sn region. These single-particle energies are crucial inputs for shell-model studies. With SPIRAL2 the evolution of thesingle-particle levels can be studied for example with transfer reactions involving medium-energybeams of nuclei along the Z=50 closed proton shell or the N=82 neutron shell.A related topic is the study of the spin-orbit splitting e.g. of the νh9/2 and the νh11/2 as well as theπg7/2 and the πg9/2 orbits. According to certain model predictions, the energy splitting of thesespin-orbit partners should decrease or even vanish far from stability for very neutron-richisotopes. To extend our knowledge on the spin-orbit splitting far beyond the doubly-magicnucleus 132Sn is a prime subject for SPIRAL2.

.

Figure 1: Experimentally known proton and neutron single-particle energies in the 132Sn region.

Shell closure far from stability and new shell gaps

As mentioned in the previous section, far from stability the classical shell gaps might vanish andnew magic numbers may show up. These new magic numbers could be those predicted by aharmonic oscillator shell model when switching off the spin-orbit coupling.

In this case, the neutron and proton numbers 40 and 70 become magic and large shell gaps arepredicted.These predictions can be tested by measuring the collectivity in nuclei around N,Z = 40, 50, 70.For example, 60Ca is one of these nuclei of interest with a magic proton number Z=20 and apossible shell closure at N=40. Other regions of interest are 78Ni with ''classical'' magic numbersor 110Zr with N=70 and Z=40.An elegant way of studying the collectivity of these nuclei and their neighbors is Coulombexcitation. With counting rates as low as ten particles per second and energies around 30-50MeV/nucleon, the position of the first excited 2+ state and the excitation strength, the B(E2)value, can be determined by measuring the γ rays from these nuclei after Coulomb excitation.This type of experiments allow for a rather quick ''mapping'' of a region of interest. As anexample, figure 2 shows the evolution of the 2+ state energy for nickel isotopes.

Figure 2: Energy of the first excited 2+ statesin even-even nickel nuclei. The shell and sub-shell closures at N=28 and N=40 are clearlyevidenced by the increase of the 2+ energies.

Similar studies can be conducted with even lower counting rates by using beta decay as a probe.The structure of 78Ni can be studied at the GANIL low-energy facility via beta decay of 78Co orafter beta-delayed neutron emission from 79Co. In the same way, 110Zr is accessible via the decaysof 110,111Y. Another proposal is to study the robustness of the Z=50 shell far from stability via thebeta decay of indium isotopes. The Z=50 shell closure creates an 8+ isomer in Cd98

48 which, if theZ=50 closure persists, should also show up in Cd130

48 .Finally, mass measurements also give crucial indications of shell behaviour from the two-neutronseparation energies.

Static nuclear moments far from stability

The measurement of the nuclear quadrupole and magnetic moments allow for a detailed study ofthe nuclear structure of the atomic nucleus. They give access to quantities like the nuclear g factoras well as the deformation of a nucleus.With radioactive beams, these studies can be performed either by using directly the high-energyradioactive beam, polarizing it by means of the tilted-foil method, or by introducing the polarizationor orientation needed for the measurements by reactions of the secondary beams. These reactions

can be of the fusion-evaporation type to study properties of isomers populated in these reactions orfragmentation reactions to study ground-state properties.Figure 3 shows a measurement recently performed at the LISE 3 separator of GANIL using theTime Differential Perturbed Angular Distribution (TDPAD) method to determine the properties ofan Iπ= 9/2+ isomer in 67Ni. The experiment allowed to determine the g-factor of this isomeric stateand to compare it to shell model calculations, constraining in this way the theoretical possibilities todescribe this nucleus. So far the TDPAD method is the only way to study g-factors of isomers inneutron-rich nuclei having life-times between 100 ns and 50 µs. Similar studies can be performedon other magic or semi-magic nuclei accessible with SPIRAL2.

Another interesting possibility for nuclear moments of exotic nuclei is to use the powerful methodof laser spectroscopy which has the bonus of providing information on the mean-square chargeradius.

Figure 3: g-factor determination of anIπ= 9/2+ isomer in 67Ni. The ionarrival time was used as time t=0 tomeasure the isomeric decay on top ofwhich the anisotropic oscillation ispresent. The oscillation pattern R(t)of the isomeric β-decay isproportional to the isomeric g-factor.

High-spin states produced with fission fragments

Superdeformed states have been identified in a large number of nuclei. However, as basically allthese studies have been conducted with stable-beam/stable-target combinations, superdeformedstates have been studied mainly in stable or proton-rich nuclei.Only a few additional studies were carried out with deep inelastic reactions to investigate high-spinstates in somewhat neutron-rich nuclei. Using neutron-rich beams to induce fusion-evaporationreactions will open completely new possibilities for the investigation of high-spin states in medium-and heavy-mass neutron-rich or stable nuclei. The excess of neutrons as compared to stable-beaminduced reactions will increase the fission barrier of the compound nuclei and thus dramaticallyincrease the survival probability. Figure 4 shows the influence of the neutron excess on the fissionbarrier. Hyperdeformation is predicted by HFB calculations around N=108. The isotopes of interest(176Er, 178Yb, 180Hf) can be produced by 130Cd, 132Sn, and 134Te induced reaction, respectively, on a48Ca target and studied with devices like EUROBALL or EXOGAM.

Figure 4: Fission barrier as a function of the mass number A and the charge number Z of the compoundnucleus. Several reactions with stable and neutron-rich beams are compared showing the influence ofthe neutron excess on the fission barrier.Figure 5 shows the potential energy for 174,176Eu as a function of the deformation parameterbeta.Another topic of interest are nuclei like 144Xe. This nucleus is predicted to have a sizeableneutron skin. High-spin states and thus rotation of such a neutron-skin nucleus has never beenobserved and might give new insight into its structure and in particular into the influence ofthe neutron skin on the rotational behavior of such nuclei.

Figure 5: Potential energy as a function of the deformation parameter for 174,176Eu as determinedby HFB calculations.

Heavy nuclei far beyond the heaviest naturally occuring element uranium have been studiedrecently in Argonne and in Jyväskylä by fusion-evaporation reactions with stable beams and targets.More neutron-rich very heavy ions may be produced by fusion reactions with radioactive fissionfragments. For example, heavy nobelium or seaborgium isotopes can be produced by 132Sn inducedreactions on 130Te and 138Ba targets. With SPIRAL2, the resulting 260No and 268Sg can be e.g.detected with VAMOS using the recoil-decay-tagging method and the γ rays can be observedwith the new EXOGAM device. Other nuclei may be produced with different beam/targetcombinations.

Search and production of super-heavy nuclei

Isotopes above proton numbers of about Z=106 are only stable or quasi-stable with respect to thestrong interaction due to the shell structure of the atomic nucleus. Without the additional bindingenergy gain from the stabilizing effect of the nuclear shells, these nuclei would not exist. Super-heavy elements have been produced with proton numbers up to Z=112 and perhaps as far as Z=116.Beyond the observation and identification of new chemical elements, one of the interests of theproduction of super-heavy isotopes is the search of new shell closures for these high- Z elements.According to different model predictions, the next neutron shell is expected at N=184. However,for the next proton shell gap, the predictions are less clear. Depending on the model, shell closuresare expected at Z=114, 120, or 126.With stable beams, it is believed to be difficult to reach these closed shells. In particular, it seems tobe difficult to produce super-heavy elements near closed shells with sufficiently low excitationenergy by means of stable beams (see figure 6). For example, the use of radioactive kryptonisotopes will allow to produce new isotopes and maybe even new elements in the vicinity of theexpected closed shells. The more neutron-rich super-heavy elements are not at all accessible withstable beams.The identification of new super-heavy elements is usually based on the observation of their decay ina sequence of α-decays. These α-decay chains allow to link the new isotope to known lighter super-heavy isotopes and thus to clearly identify the super-heavy nuclide produced. However, for themore neutron-rich super-heavy isotopes, these links to known isotopes do no longer exist, becausethe properties of the lighter decay products are not known. Therefore, it is of crucial importance tostudy the reaction mechanism to produce neutron-rich nuclei of elements between Z=100 andZ=110 as well as their decay properties. These isotopes can be produced with reasonable countingrates with neutron-rich fission fragments from SPIRAL2. The energies needed are around theCoulomb barrier.

Figure 6: Excitation energy of compound nuclei for the production of super-heavy isotopes as afunction of the isotopes produced with stable or radioactive beams. Shown are also the knownisotopes for nobelium and element 112. The black circles are stable projectiles used for theproduction of the super-heavy isotopes. In red and blue are the radioactive isotopes used. Inaddition, predicted half-lives are shown. The bottom part indicates the stable (black dots) andknown radioactive nuclei that can be used as they will be produced with high intensity. The x axisgives the neutron number of the projectile as well as of the compound nucleus (courtesy of S.Hofmann).

Nuclear astrophysics

To understand stellar evolution and the production of the elements in the universe, extensive modelcalculations are used to describe and simulate the different processes occurring in the stars.Especially for violent processes like supernovae explosions or X-ray bursts, mainly properties ofunstable nuclei are the most important inputs to the models. In neutron-rich stellar environments,the rapid-neutron-capture process produces heavy elements by a sequence of neutron captures andnuclear beta decays. To correctly model this r-process, the model inputs needed are masses of veryneutron-rich nuclei, their beta-decay half-lives and their neutron-capture cross sections. However,up to now, these properties are only known for a few isotopes involved in the r process. Neutron-rich fission fragments from SPIRAL2 will allow to perform measurements of half-lives and massesfor some of the key nuclei (see figure 7).

Figure 7: Left-hand side: Chart of nuclei of measured masses in the r-process region accessible withSPIRAL beams. The yellow part shows the region of nuclei produced by SPIRAL2 withphotofission. In pink is indicated schematically the path of the r-process. The black squares indicateknown masses where the size of the square is proportional to the uncertainty of the experimentalmass. Right-hand side: Chart of known half-lives with the size of the squares proportional to theirerror-bars.

Neutron-capture cross sections are usually estimated by means of Hauser-Feshbach calculations.These calculations assume that a statistical picture is applicable for the capture process. However,close to magic numbers, and especially far from stability, the level density becomes so low that thestatistical picture is no longer valid. Therefore, measurements are needed to determine these capturecross sections. A possibility is to measure cross-sections for neutron transfer via a (d,p) reactionwith neutron-rich radioactive beams and to determine from these measurements the capture crosssection for neutrons. This cross-section is known to be strongly influenced by the low lying strengthof the giant dipole resonance and therefore coincidence measurements of fl rays are important asthey provide information on the oscillation modes of the less bound neutrons. Regions of interestfor these measurements are close to the magic shells at Z=28, N=50, and N=82.

Thermodynamics of the nucleus

In the following sections, we will address measurements that will be performed with a combinationof stable and radioactive beams. They deal with thermodynamical properties of the atomic nucleus.Starting from less perturbative reactions, we will go to more and more violent collisions whichpump more and more excitation energy into the nucleus.

Level densities and entropyThe level density is a basic ingredient of the nuclear equation of state. It allows to localize theopening of decay channels, to study correlation phenomena in the nucleonic movement, and toobserve phase transitions. In addition, the level density is an important input for cross sectioncalculations and its knowledge is essential for our understanding of astrophysical processes. Atpresent, the level density has only been studied in nuclei close to the line of stability at low

excitation energies. Recently, a new methodology has been proposed to measure level densities forstable nuclei. The measurements are based on particle-α coincidences which allow to determine thelevel density for neutron-rich nuclei up to excitation energies of about 10 MeV. At the same time,phenomena like the quenching of pairing effects, the temperature dependence of the pygmyresonance, or the disappearance of shell effects can be studied. Using radioactive nuclei fromSPIRAL2, these measurements will be extended to exotic neutron- rich nuclei around mass A=150.

Coulomb instabilities and limiting temperatureThe limiting temperature a nucleus can stand is linked to a decrease of the level density at highexcitation energies and its measurement helps to constrain the nuclear equation of state at finitetemperatures in the vicinity of the saturation density. For stable nuclei, the limiting temperatures arerather high and, to investigate this parameter, one needs to control the energy deposited, the degreeof equilibration, and the reaction mechanism.

Figure 8: The limiting temperature is plotted on a graph of the temperature versus the mass of thecompound system. By changing the incident energy and the projectile/target combination from aproton-rich to a neutron-rich combination, different regions below and above the limitingtemperature can be reached.

In the vicinity of the proton drip line, Coulomb instabilities are predicted to decrease drastically thelimiting temperature (see figure 8). The proposal is to study an isotopic chain of compound nucleiproduced via fusion-evaporation reactions from the proton-rich side (SPIRAL1) to the neutron-richside (SPIRAL2) by precisely measuring the deexcitation pattern of particle emission in coincidencewith evaporation residues in an excitation energy range between 1 MeV/nucleon and 3MeV/nucleon. Additionally, these studies allow to investigate the influence of the Coulombinteraction on the expansion of a nuclear system initially strongly compressed.These studies require projectiles of e.g. 114Xe to 145Xe at energies of 5 MeV/nucleon to 30MeV/nucleon with intensities of about 106 particles per second impinging on calcium targets. Othersystems to be studied are based on beams of 74-96Kr impinging on iron targets.

Reaction mechanism and thermodynamical equilibriumThe degree of thermal equilibration of a nuclear system can be studied as a function of thedeposited energy by detecting particles emitted in the reaction. These studies, performed over wideranges of isospin, provide information necessary for investigations of the thermodynamics of

nuclear systems at high excitation energies in order to identify an equilibrated source of particleemission and thus to control the reaction mechanism.Using projectile/target combinations with different isospin, the mapping of the N/Z ratio of lightparticles as a function of the rapidity allows then to determine the degree of equilibration of anuclear system.This technique has been employed in experiments at the FOPI detector of GSI, where a large part ofthe energy available is collective. At lower incident energy, a large part of this energy should beavailable to thermalize the system. In order to perform this kind of studies, a detector capable toidentify fragments in A and Z is necessary. As the degree of equilibration depends on the mass ofthe system, light and heavy systems have to be investigated. Some of the systems (e.g. 83-106Nb on54,58Fe or 114-145Xe on 112,124Sn) can be studied at SPIRAL1 and SPIRAL2.

Transport properties and symmetry energiesSophisticated transport models predict large-amplitude fluctuations of the matter density and of theisospin content in the interaction region of binary collisions at the Fermi energy.These fluctuations can be observed by measuring isotopic distributions of fragments at mid-rapiditybetween the quasi-projectile and the quasi-target. A comparison to theory allows to link thedistribution to the dissipation properties of the nuclear transport and to the density dependence ofthe symmetry coefficient of the equation of state of asymmetric nuclear matter. To control thereaction mechanism and the degree of dissipation, it is necessary to detect coincidences between thequasi-projectile and the intermediate mass fragments which have to be identified in mass andcharge. These studies can be performed with similar beams as those mentioned in the previousparagraph.

Multifragmentation and the liquid-gas phase transitionThe observation of a negative heat capacity via the measurement of fluctuation in the reaction Qvalue has allowed recently to demonstrate that multifragmentation corresponds to a liquid-gas phasetransition. The same analysis performed with exotic nuclei will allow to study the phase diagram asa function of isospin. The observation of fossile signals of the spinodal decomposition in the chargecorrelations between fragments suggests that the phase transition is driven by density fluctuationsamplified by the mechanical instabilities of the spinodal region.Theory predicts that, for neutron-rich systems, the phase transition is due to a chemical spinodalzone which leads to a fractionation of isospin: the heavy fragments are more proton rich, whereasthe lighter fragments are neutron rich. Again, an unambigous determination of the fragments in Aand Z is crucial to perform these kinds of studies. Incident energies of the order of 30 MeV/nucleonare necessary to reach the threshold for multi-fragmentation. It is of interest to study exotic isotopesof the same systems for which the phase transition has been observed. Therefore, beams of 114-145

Xe on tin targets as well as gold-on-gold reactions (176-205 Au + 197Au) are required.

SummaryThe subjects mentioned here represent only a few selected topics of nuclear physics intereststhat can be addressed using beams of neutron-rich fission fragments. The presented casesshow that a wide variety of problems in nuclear physics and nuclear astrophysics can thus beaddressed. Therefore, low- and medium-energy beams from SPIRAL2 up to some 50 MeV/A

will open new exciting possibilities in different fields of nuclear physics and applied nuclearphysics.

Production of heavy neutron-rich beams

Fast neutron induced fission

The aim of the SPIRAL-II project is to establish the feasibility of producing accelerated beams ofneutron-rich radioactive beams with the existing radioactive beam facility SPIRAL at GANIL. Theneutron-rich radioactive nuclides are to be produced by fissioning a heavy nuclide, such as 238U.The technique originally proposed for SPIRAL-II is the use of energetic neutrons to induce fissionof depleted uranium. The neutrons are generated by the break-up of deuterons in a thick target, theso called converter, of sufficient thickness to prevent charged-particles to escape. The energeticforward-going neutrons impinge on a thick production target of fissionable material. The resultingfission products accumulate in the target, diffuse to the surface from which they evaporate, areionised, mass-selected and finally post-accelerated. This method has several advantages. The highlyactivated converter can be kept at low temperature without affecting the neutron flux. The target isbombarded by neutral projectiles losing energy only by useful nuclear interactions and having ahigh penetrating power allowing very thick targets.

One of the main objectives of the R&D program (for details see ref [1]) was to determine theintensity and energy of the primary deuteron beam giving the best yields of radioactive nuclides ofinterest for radioactive beams while taking into account beam power evacuation and safe operationof the facility. The approach has consisted in carrying out simulations with various codes availableor developed by our different task groups and performing a number of key experiments to validatethe simulations. In this way, confidence is gained about the predictive power of the codes forsituations where experiments could not be set up within the allocated time for the study.The concept of using neutrons generated by deuteron break-up implies a study of production yields,energy spectrum and angular distributions of neutrons in converters made of various materials andas a function of deuteron energy. Experiments were performed at IPN-Orsay, KVI-Groningen andSaturne at Saclay. They explored a range between 14 and 200 MeV deuteron energy. The mainfeatures of neutron spectra are listed below

At forward angles, the energy distribution has a broad peak centred at about 0.4 times the deuteronenergy. The angle of emission becomes narrower with increasing energy. For 100 MeV deuterons,the energy width (FWHM) of the neutron spectrum is about 30 MeV and the FWHM opening angleof the cone of emission is about 10 degrees.

There is a rather isotropic distribution of neutrons of a few MeV due to evaporation in fusionreaction.

The angular distributions and energy spectra are in fair agreement with calculations with anextended version of the Serber model and with the LAHET code. The Serber model reproduces thedistributions of high-energy neutrons but not of the low-energy neutrons since evaporation is notimplemented in the code. LAHET reproduces the low energy neutron spectrum while it tends toslightly underestimate (less than a factor 2) the neutron distributions at very forward angles.

A strong increase in neutron production is observed between 14 and 100 MeV deuteron energy. It ismuch less pronounced between 100 and 200 MeV. Among converters tested Be is slightly moreproductive than C. It has, however, disadvantages related to its physical and chemical properties.

Experimental angular distributions of neutrons for 80 and 160 MeV deuterons incident on a thickBe target (left). Neutron yield at 0° as a function of the incident deuteron energy for a Beconverter(right).

Cross sections for n-rich nuclei in neutron-induced fissionAnother step is the measurement of fission cross sections for n-rich nuclei at intermediate neutronenergies. Magnitudes and widths of distributions for neutrons of average energy 20 MeV generatedby the 50 MeV deuteron beam on a 238U target are similar to those obtained with 25 MeV protons.Yet, for a given element, they are shifted towards larger mass by about 2 mass units. In that respect,they are similar to the distributions by thermal n-induced fission on 235U. However, fast neutronsproduce a much wider distribution with higher cross sections at very asymmetric mass splits(A=80,160) and in addition, the dip in the symmetric region (A=120) is almost filled. These resultshave been used to extend a model originally designed for fission induced by intermediate energyprotons.Comparison with data for 2.5 MeV neutrons shows that about 1.4 neutron is lost when the neutronenergy is increased from 2.5 MeV to an average of 20 MeV. Due to increased excitation energy ofthe compound nucleus, high energy favours production of less n-rich fragments. On the other hand,the total fission cross section increases up to about 40 MeV neutron energy and neutron productionfurther increases. This implies the existence of a certain optimum for producing neutron rich nuclei.

Studies with a system formed with converter and targetIt is also necessary to consider the geometry of the converter + target assembly. The target mustintercept a large fraction of the neutrons. Thus, it must be close enough to the converter to subtenda high solid angle. The target temperature must be high enough to allow fast and efficient release ofthe fission products. These studies have been performed with devices designed and constructed atthe IPN-Orsay referred as PARRNe1 and PARRNe2 [2]. They both include a converter and a target.Various materials for the converter and two different targets, a high temperature porous UCx and aliquid uranium target, have been tested.

In a recent experiment at PARRNE2 with the fast release UCx target (December 2000) a MK5plasma source similar to the one at CERN-ISOLDE has been used. PARRNE2 indeed is able todeliver most of the elements available at ISOLDE. Among the very encouraging results is theseparation of the double magic 132Sn produced, extracted from the target, singly 1+ ionised andmass separated with a rate of 3.5 105 / s. Extrapolation (from measurements done at differentenergies on noble gases) at 100 MeV and 500 µAe let expect for 1.7 109 single-charged ions / s of132Sn delivered for charge breeding and post acceleration. The rates for elements extracted andionised as singly charged 1+ ions are shown on the figure below.

Diffusion depends critically on the structure of the target material and on the temperature. At IPN-Orsay, R&D has been devoted to UCx targets following the technique elaborated at ISOLDE and tomolten uranium targets. The UCx and molten target are complementary. The higher temperatureand shorter path to the surface in the grains of the porous UCx target favour fast diffusion (Trelease =5 s for Xe). The molten uranium target has a larger amount of uranium but slower release (Trelease=112 s for Xe). The former is more efficient for short-lived activities, while the latter is more suitedfor longer-lived activities up to half-lives of 30s. Further research is needed in order to find theoptimum material structure, such as composition and density, for the targets and coupling to the ionsource. These properties play a major role in the diffusion and thermal behaviour of the target.

According to several experimental observations the deuteron energy of 100MeV is near the bestvalue. In short, this results from the energy dependence of neutron yield, neutron energy spectrumand angular distribution the energy dependence of total fission cross section, which reaches amaximum at 40 MeV neutron energy (i.e. 100 MeV deuteron) and the energy dependence ofnuclide cross-sections. For a fixed number of total fission, nuclei of asymmetric region are similarlyproduced at lower energy while the others at higher energy. An energy around 100 MeV appears tobe a good compromise.The LAHET and FICNeR codes also show that the number of fissions per kW of accelerateddeuteron beam first strongly increases with energy but near 100 MeV saturates.

Power load on target and beam intensity limitThe maximum beam intensity which can be withstood by the converter or target define themaximum production rates achievable at the best deuteron energy:The converter can accept a beam intensity of 350 µAe at 100 MeV if the beam is defocalised to a 2cm radius. The limit is actually due to cooling of the outer surface of the C-converter. With a higherheat exchange coefficient (but within a factor of 2 higher) than presently the GANIL standard, itshould be possible to accept 500 µAe on the converter. An alternative solution would be a fastrotating converter.The conventional method is to let the beam to impinge directly on the target. In this case, it is betterto let the beam to exit the target. This avoids unnecessary heating by the ions close to the end oftheir range (the Bragg peak). Nevertheless, only a 50 µAe beam of 2 cm radius is possible on thistarget. The major drawback is certainly the lack of temperature control else than via the beamintensity.

Yields have been obtained with the LAHET+MCNP+CINDER code for radioactive beamintensities consistent with temperatures which the target can stand and an energy of 100 MeV. The350 µAe beam on the converter induces 6 1012 fissions / s in the target while this is 1.2 1013

without converter with 50 µAe beam. However, yields of the most neutron-rich nuclei are higherwith the converter, owing to the lower projectile energy (neutrons of 40 MeV average energy) andthe formation of a more n-rich compound nucleus (no proton is captured).

Ionisation

The atoms diffused out of the thick target must be ionised. The most recent PARRNe2 experiment,with a MK5 plasma source designed at ISOLDE, was promising. However, depending on theelement to be ionised it is necessary to use several sources to guarantee for highest efficiency and/orselectivity. So far considered are a ECR-source for gases and volatile elements, e.g. MONO1001 atGANIL, a surface ionisation source for alkaline and earth-alkaline elements and a more universalplasma source with hot transfer line, e.g. the ISOLDE MK5 source. These sources could haveefficiencies close to 100% for rare gases and alkalines, and between 10% and 50% for otherelements. Moreover, a laser ion-source is well suited for specific cases where especially high purityis compulsory.

Charge breeding is necessary in order to accelerate the ions to final energies above the Coulombbarriers. Typically, the charge state must be increased up to at least one third of Z. Efficiencies arepresently between 3% and 12% for the most probable charge, depending on the elements.

Safety considerationsThere is, to date, no facility comparable to the design goals of SPIRAL-II. Therefore, radiationdoses and production of contaminants have been estimated by using the LAHET-MCNP-CINDERcode at GANIL. This has been carried out for different configurations, converter material anddeuteron energies. Considering the radiation doses (proportional to the beam intensity) for the sameproduction of nuclei of interest, we arrived to the following conclusions:Alpha-activities one year after the beam stopping are produced in large amounts at the highestenergies since reaction channels for actinides is opened for more nuclei at higher energy. The α-activity is higher for deuteron beam on a UCx target in the direct method than with a converter.Tritium production is larger with light-Z converters (C, Be or Li) than without. Carbon is thelowest producer of tritium among them.

For a fixed number of fission/s, the radiation doses during operation and most residual activitiesafter a long term shut down are weakly depending on the energy. The only exception is for tritiumproduction, which increases monotonically by a factor of 9 between 50 MeV and 200 MeV.Nevertheless, this increase is a factor of 2.5 from 50 MeV to 100 MeV. Production of alpha-activities is clearly the lowest near 100 MeV. In conclusion, the energy of the deuteron beam forsafe operation, taking into account the production of α emitters, is 80-100MeV. This is the samerange as it was determined to be best for production of neutron-rich nuclides.Moreover measurements have been performed. The attenuation length of neutrons in concrete hasbeen measured at GANIL in order to better estimate the amount of concrete shielding for neutronsneeded during irradiation. Activation of air in the target area has been measured at Louvain-La-Neuve to validate the safety codes. Finally, a method has been developed and tested in order tomeasure the amount of tritium escaping from the target by diffusion.

Photofission

It has recently appeared that photofission could be an alternative to n-induced fission. This workinggroup has therefore initiated a study of photofission induced by Bremsstrahlung generated byelectrons. The respective merits and technological challenges of these two methods will beevaluated in order to made the best choice for SPIRAL-II.

With an electron driver, electron interaction with matter will radiate Bremsstrahlung photons insidethe target. Fission will then be induced by those photons exciting the Giant Dipolar Resonance(GDR) of the nucleus at the right energy. This well-known process is called photofission.The GDR cross section for 238U is shown in the figure below on the left. A maximum fissionprobability of 160 mb is obtained for photons having energy around 15 MeV. At that energy, thephotoelectric and the Compton and Rayleigh scattering cross sections are starting to fall off rapidlyso the main contributions to gamma absorption are e+e- pair production and the photonuclearreactions (γ,f), (γ,n) and (γ,2n). Although the absolute fission cross section is rather small(compared to normal fission with neutrons), its contribution is not negligible as even a pairproduction reaction may in a thick target eventually lead to a fission through the resulting photonproduced. In the same manner the neutrons produced by (γ,n) and (γ,2n) reactions as well as the (γ,f)itself can also induce fission, this time by the regular (n,f) high cross section (0.5 barn for fastneutrons). Therefore, in a thick target, photofission may be a rather interesting way of creatingradioactive fission fragments.

0

50

100

150

200

0 5 10 15 20 25 30

Photon Energy (MeV)

σ σσσ (m

barn

)

U238 Exp

U238Theoretical fit

Photofission cross-section for 238U above (left and right)and rate of fission per electron according to incident energy of the electrons (right).

U238 Target, E e = 50 MeV, t = 4 mm

1.E-03

1.E-02

1.E-01

1.E+00

1.E+01

0 10 20 30 40 50

Photon Energy (MeV)

(Gam

ma

/ inc

iden

t ele

ctro

n) Total produced photonsOutput Photons

Photon energy distribution by Bremsstrahlung in 238U

Unfortunately, no efficient monochromatic sources of 15 MeV photons are available. The mostcommon way for producing high gamma fluxes is the Bremsstrahlung radiated by passage ofelectrons through matter. This process has a cross section rising linearly with energy. It willdominate the ionization process above a critical energy (around 20 MeV). But the resultingBremsstrahlung spectrum is widely spread in energy from zero up to the full initial electron energy(figure above). Although each single electron may ultimately produce as high as 20 photons, only asmall fraction of it (0.5 to 0.7 gamma per e-) are "useful" photons lying in the GDR range (15±5MeV).A simple calculation including the main electron interactions (Bremsstrahlung and ionization) andthe main nuclear reactions of interest (pair production and fission cross section) in a thick depleteduranium target can give the expected number of fission per incident electron. In the figure below,the number of fissions produced by the (γ,f) reaction is plotted as a function of the electron energy.This result is a complete Monte Carlo calculation performed with a MCNP code offering alsophotonuclear capability (full electron, photon and neutron transport). The obtained result is inaccordance with the simple analytic calculations. For comparison, fission production is also givenwhen using a tungsten converter (5 mm thick) in front of the 238U target. It appears that when usinga converter in the electron driver option, less than 30% of the beam power is lost inside theconverter (in contrast to the d driver option). In the direct method one will produce about 25% morefission per electron (and probably more when taking in account neutron induced fission). Fissionproduction is almost linear above a threshold energy of 10 MeV. High production is obtained above40 MeV.As the SPIRAL II project is aiming at 1013 fissions/s, the required beam intensity should be 500 µAfor an electron energy of 45 MeV directly on the target.

In order to compare rapid neutron induced fission and photofission, measurements of Kr and Xeisotopic distributions produced by photofission and diffused out of a thick UCx target has beenperformed using the same PARRNe1 device in the same conditions that with deuterons beams andwith the same target. The measurements have been done with a 4 mm W converter in differentposition (8 mm from the target, 4 mm from the target) and one measurement without W converter.Comparison with the 80 MeV deuteron induced fission measurements are presented below.

The results obtained are well understood taking into account the percentage of photons between 11and 17 MeV emitted in the cone subtended by the target that is the solid angle.Finally, the above figure summarizes some comparison of production yields normalized to 1013

fissions for 100 MeV deuterons (LAHET calculations) and 50 MeV electrons based of the K-HSchmidt estimations.

In the case of the electron driver, much work remains to be done in the design of the converter andof the UCx target specially to resolve problems related to the beam power dissipation in the target.This work is still in progress and constitute one major point of the success of inducedphotofission with an electron beam.

The deuteron driver for the SPIRAL-II project

Three options for the accelerator providing the deuteron beam for the neutron production in theSPIRAL-II project have been studied

• the actual GANIL cyclotrons• the SARA booster-cyclotron with a new injector• a new cyclotron.

In the present chapter the specifications of the deuteron accelerator and the different options forsuch an accelerator at GANIL are discussed.

Accelerator typeFrom a technical point-of-view both a cyclotron and a linear accelerator are suitable to produce the80 MeV deuteron beam needed. For cyclotrons the presently achieved maximum beam intensitiesare about 1 mA of 30 MeV protons in commercial compact cyclotrons used for isotope productionand 2 mA of 72 (and 590) MeV protons in the separated sector cyclotrons at PSI (Switzerland).For linear accelerators the achievable currents are significantly higher. However, at the low energy(80 MeV) needed for the present application and assuming that the required intensity can beprovided by a cyclotron linear accelerators are much less cost-effective. They have a largerfootprint, thus requiring a higher investment for infrastructure (building etc.), while also theaccelerator itself is far more expensive than the equivalent cyclotron. Finally, if CW operation isrequired only a linear accelerator with superconducting RF cavities can be used, which furtherincreases the investment.It is concluded that for the SPIRAL-II project the use of a cyclotron is the best option: it permits toachieve the objectives of the project at the lowest costs. However, as a consequence of this choice, asignificant increase in intensity in the framework of a future upgrade will require large additionalinvestments.

Cyclotron characteristicsRadiation safety is a very important issue when accelerating deuterons to an energy of 80 MeV withan intensity up to 0.5 mA. Beam losses around 1 % already cause strong activation of the parts ofthe cyclotron on which the lost beam impinges. Furthermore a large flux of high energy neutrons isproduced, which will significantly activate the cyclotron and its surroundings. This flux of neutronsis increased even further because of the low threshold for neutron production in the d(d,n)3Hereaction on lost deuterons that have been implanted in various parts of the cyclotron. Furthermoretritium is produced via the reaction d(d,p)3H. This may lead to additional radiation safety problems.It is thus essential that a very high transmission (≥ 99.9 %) is achieved, in particular for the laterstage of acceleration and for the extraction. It is the prime factor in assessing the suitability of thethree possible systems studied:

The actual GANIL cyclotronsThe actual GANIL cyclotrons can accelerate deuterons to energies in the range from 12 to 26 MeVand probably around 80 MeV. With the injector cyclotron and the first separated sector cyclotronthe energy range 12 to 26 MeV can covered without modifications of the accelerators. The fissionyield at 26 MeV is at least an order of magnitude lower than at 80 MeV, taking into considerationthe energy dependence of the neutron production and fission cross sections.Around 80 MeV it seems possible to accelerate deuterons using the complete chain of injectorcyclotron and two separated sector cyclotrons. In the injector cyclotron and the first separated sectorcyclotron molecular ions would be accelerated, which would be stripped into two deuterons oninjection in the second separated sector cyclotron. As the neutron production by deuterons at thisenergy is at least an order of magnitude higher, radiation safety problems will impose much moreserious constraints, connected to both the shielding and the activation of components.The feasibility of 80 MeV deuteron beams as well as the radiation safety and technical constraintson the beam intensity are subject to further study.

The SARA booster-cyclotron with a new injectorThe SARA booster-cyclotron is a separated sector cyclotron, which is presently beingdecommissioned at the Institut des Sciences Nucléaires in Grenoble (France). It can acceleratedeuterons to a maximum energy of 72 MeV, slightly lower than the value aimed at. It has an energygain of a factor 5, so that an injector (cyclotron) delivering 14.5 MeV deuterons is needed. Thematching conditions between the injector and booster cyclotron strongly constrain thecharacteristics of the injector. Consequently the injector will have to be specifically developped.Furthermore the injection, RF-system and extraction of the booster will have to be reconstructed tomeet the requirements of the high intensity operation.The requirements on the extraction efficiency can only be met by the use of stripping extraction ofnegative ions. These negative ions thus have to be extracted from the injector by a standardextraction system using an electrostatic deflector. Taking into consideration the lower yield andenergy of neutrons produced at 14.5 MeV as compared to 80 MeV the extraction efficiency of theinjector should be at least 95 %, which is close to the limit of feasibility.

A new cyclotronAn analysis of the existing cyclotrons delivering high intensity proton beams (≥ 1 mA) at a fixedenergy shows two possible schemes to achieve the required transmission:A large, low-field separated sector cyclotron with high energy gain per turn and 'classical' extraction(e.g. PSI injector II: 2mA 72 MeV protons). Cyclotrons of this type have been developpedexclusively for research institutions. The high transmission is obtained by maximizing the radialdistance between subsequent turns in the machine, so that an electrostatic septum, which bendsaway the last turn, can be inserted inbetween two turns.A cyclotron for 80 MeV deuterons based on this approach would have an extraction radius of 3.5m.A compact cyclotron accelerating negative ions with stripping extraction (e.g. IBA (Belgium)CYCLONE30 and EBBCO (Canada): TR30, both delivering 1 mA 30 MeV protons). Around 20cyclotrons of this type are used routinely for isotope production in an industrial environment. The

efficiency of the extraction process is about 100 %; the overall transmission is determined by beamlosses during acceleration caused by the magnetic field and the interaction with the residual gas.A cyclotron based on these examples would have an extraction radius of about 1.6 m. For thepresent application the superior beam quality of the first, far more expensive scheme is notrequired. Furthermore, the use of stripping extraction makes it possible to vary the energy over arange of roughly a factor two (40 – 80 MeV) by changing the radius at which the stripper foil islocated.

Analysis and recommendationThe use of the present GANIL-cyclotrons requires by far the lowest investment. The investment fora dedicated accelerator (compact cyclotron or SARA booster + injector) is estimated to be of thesame order of magnitude (10 – 12 MEuro excluding infrastructure).

The installation of a dedicated accelerator will make the operation of the SPIRAL-facility (and ofthe GANIL-facility in general) more flexible, efficient and versatile:

• development work on beams with SPIRAL-II can proceed while beams from SPIRAL-I orthe present GANIL-facility are delivered for experiments.

• operation of the SPIRAL-facility is decoupled from that of the present GANIL-cyclotrons.• the intensity of the radioactive beams attainable is at least an order of magnitude higher,

thus significantly extending the range of feasible experiments.• the possibility to inject beams from SPIRAL-II into the present GANIL-cyclotrons remains

open.Simplicity of both concept and operation is an important asset for the deuteron accelerator inthe SPIRAL-II project. The compact cyclotron accelerating negative ions has by far the bestscore on this aspect.The development and construction connected to the use of the present GANIL-cyclotrons or theSARA-booster will require a large effort from the accelerator staff of GANIL (or other possiblepartners in the project). The development and construction of the compact cyclotron, based onexisting industrial designs, is suitable for contracting to industry.It is thus concluded that a dedicated compact cyclotron accelerating negative ions is theoptimal solution for the deuteron option for a short term in the SPIRAL phase II -project.


General LayoutThe accelerator layout is shown below and is quite similar to the MACSE project [3] The injectorcomprises a 100 keV gun and a short cryomodule containing a single superconducting cavity. At theinjector exit, the electron beam is already well relativistic. Then the beam runs in a longcryomodule through four SCRF cavities bringing the electrons to the final energy of approximately45 MeV. A bending magnet followed by a transport beam line will get the beam at the right placeand shape onto the target.

InjectorInjector



beambeam dump dump

analysisanalysis

targettarget

beambeamdumpdump


deviationdeviation



12 m


GunThe gun uses a standard thermoionic cathode sited on a 100 kV platform. The power supply candeliver up to 3 mA in current. In order to bunch the beam at the proper frequency of 1.5 GHz, aspecific line consisting of two rectangular RF cavities (chopper) are used to cut a 60° phase out ofthe DC beam delivered from the cathode. Another cylindrical cavity (buncher) is used to bunch thebeam down to a 10° phase at the entrance of the capture cavity.An attractive alternative method would be to set up a gridded cathode similar to the one used inInductive Output Tubes (IOT). In that way, the beam could be bunched right from the startingemission point that suppresses the need of the rather cumbersome bunching line. This newtechnique will be tested at CEA/Saclay in 2001 and if successful, will be implemented on theSPIRAL II machine.A collimator can be placed after the gun in order to control the beam emittance. This could be ofimportance for safety and radiation issues (see below) as undesired beam tails can be quite easilysuppressed at that stage without creating too much radiation.

Capture CavityThe capture cavity is a superconducting cavity with a reduced beta value due to the not fullyrelativistic beam coming out from the gun. Geometry has to be optimized to properly capture thebeam without degrading too much the longitudinal emittance. The figure below shows somesimulations where different cavity shapes have been studied.

0

1

2

3

4

5

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8z (m)

ener

gy (M

eV)

ββββ=1 cav

hybrid cavββββ=0.85 cav

@Bpk=50mT

0

1

2

3

4

5

6

7

8

-180 -150 -120 -90 -60 -30 0 30 60 90 120 150 180

phase z=0 (deg)

ener

gy (M

eV)

hybrid cav

ββββ=0.85 cav

ββββ=1 cav

@Bpk=80mT

Output energy as a function of position and phase for three different cavity shapes. A β = 1 cavity isclearly not adequate to use as a capture cavity.

The output dispersion of the beam shown on the figure below is simulated for two different cavityshapes assuming an input beam with 200 eV and 30° extension in phase. It can be noticed that theoutput beam will feature less than 40 keV and 5° in the longitudinal phase space.

-60

-50

-40

-30

-20

-10

0

10

20

-20 -15 -10 -5 0 5 10 15 20

bunch entrée

sortie beta=0.85

sortie hybridcav

∆ϕ∆ϕ∆ϕ∆ϕ(deg)

∆∆∆∆E (keV)1000 part. shot Bpk=50mT -> 4.7MeV, 4.3MeV

Longitudinal dispersion at the output of the capture cavity for two different shapes.

SCRF CavitiesThe four SCRF cavities are standard β=1 elliptical cavities working at the resonance frequency of1.5 GHz. The characteristics of these cavities are summarized in Table I. The important thing topoint out is the high accelerating field operation (19 MV/m) and the very low bandwidth (40 Hz)due to the low beam current.The cavity will use an integrated helium tank vessel and a specific tuner will have to be developed.A great care should be taken as regard to the low bandwidth. Therefore, any improvement made ontuners developed in the frame of the CEA/Saclay-IPN/Orsay collaboration for SCRF cavities couldbe implemented in SPIRAL II. For example, the use of piezoelectric tuners if demonstrated wouldbe of great benefit.

SPIRAL II SCRF CAVITY

Frequency 1 500.000000 M Hz Beam Power 4.756 kW

β 1.000 Dissipated Power 9.058 W

λ 0.200 m Incident Power 4.767 kW(β∗λ /2) 0.100 m Incident Q 3.80E+07(r/Q )/cell 50 Ω Zero Current Voltage V0 1.90E+07 V

Slope X 1.90E+10 ΩNum ber of cells 5 V 9.52E+06 V

Bpeak 80 (m T)Bpk / Eacc 4.2

Eacc 19.05 (M V/m )Q 0 2.00E+10 All External Q 's 1.00E+11

Length 0.500 m Loaded Q L 3.79E+07(R/Q ) 250 Ω Bandwidth (2.∆f) 40 Hz

Stored Energy 19.222 J α inc 0.997724

α cavity 0.001897

M axim um Voltage 9.52E+06 V α ext 0.000379Beam Current 5.00E-04 A

Phase 2.000 degrees Reflected Power 1.72E-04 WActual Voltage 9.51E+06 V Cavity Losses 9.058 WEnergy G ain 9.512 M eV Transm itted Power 1.812 W

Detuning Angle -1.992 degreesCavity Frequency 1 500.000 M HzFrequency Shift -0.69 Hz

Table I – Typical SCRF cavity characteristics.

CryomodulesTwo separate cryomodules are needed. The first will house the single capture cavity and the secondone the four SCRF β=1 cavities. These cryomodules are similar to the one used on the MACSE testbed but will be modified to take in account the fact that the helium vessel is suppressed, each cavityhaving an individual helium tank. Schematic drawings of the largest cryomodule are shown in thenext figures.

Table SupportEnceinte hélium

SouffletCoupleur de Puissance

Enceinte à Vide

Ecran AzoteCavité

Longitudinal view of the cryomodule.

The assembly of cavities and couplers has to be done inside a class 100 clean room in order toavoid dust contamination. The assembly can be done outside the clean room provided thecavity/coupler ensemble is leak tight and sealed. The insertion in the cryomodule of the full 4-cavityassembly has to be done from the side.

Front view of the cryomodule. The cavity is laid on a cooled table in a cradle shape structure on thebottom side of the helium vessel.

Power CouplerThe power coupler is an important (and weak) element bringing the RF power from roomtemperature down to the cold cavity. It should be stressed that even though the power level is ratherlow (5 kW), this item should be developed and fully tested during the TDS phase. If not, the overallschedule would shift as the power coupler is on the critical path. Moreover, in view of a futureupgrade, it would be very interesting to check whether the power coupler could be designed towithstand a CW power of 100 kW. Some basic parameters for the two options (5 kW & 100 kW)are described in Table II. Basically the external diameter should be higher for the high poweroperation to avoid the multipacting bands. The mechanical impact on the cryomodule mounting hasalso to be thoroughly analyzed. The TDS work should determine if a high power coupler could besafely implemented at that stage or not.

Waveguide to Coax transition

Window

Cavity Cavity flange

Schematic design of the coaxial coupler for SPIRAL II.

Parameter 5 kW 100 kWExternal Diameter (mm) 41.3 (standard 1"5/8) 80Impedance (ohms) 50Internal Diameter (mm) 17.9 35First Multipactor Barrier(kW)

125 1765

Total Dissipated power at300K (W/m)

25.7 514 265

Inner Conductor DissipatedPower at 300K (W/m)

18 358 185

Outer Conductor DissipatedPower (W/m)

7.8 156 80

Inner Conductor MaximumField (kV/cm)

0.95 4.23 2.18

Outer Conductor MaximumField (kV/cm)

0.41 1.84 0.95

Dielectric Losses in theWindow (W)

1.4 28 28

Inner Conductor Lossescavity side (W)

5.4 107 55

Thermal Gradient at theWindow (K)

3.5 68.7 42

Table II –Coaxial power coupler parameters for both the 5 kW and 100 kW options.

RF SourceThe RF power is a 5 kW CW klystron TH2466 from Thalès (formerly Thomson) company. Acomplete RF power source includes :

• The klystron tube (5 kW CW)• A power supply (11 kV, 1.2 A)• A waveguide WR650 circulator (5 kW any phase)• A water power load WR650 (5 kW CW)• Miscellaneous waveguide components, couplers, elbows, transitions, etc…

The RF tube could be installed right close to the cryomodule minimizing the length of waveguidesand reducing the RF losses.

Low Level RFSCRF cavities have to be regulated in frequency, field amplitude and phase. A precise regulation isrequired to maintain the cavity at the right frequency and the accelerating field at the right value andphase. The frequency is controlled using the cavity tuning system to better than 1Hz. It is generallya rather low speed (typically one second) when using the stepping motor. A high gain feedback loopfor cavity phase and amplitude is necessary to compensate for mechanical vibration andmicrophonics. It should be possible to obtain with a properly designed phase lock loop a fieldcontrol better than 1% and a phase control better than 0.1 degree. Each cavity has to be driven by asingle RF power source. This is very important to guarantee the phase and amplitude control.Beam TransportThe beam transport is rather simple taking in account the fact that a 45 MeV electron beam ishighly relativistic (γ ∼ 90). In particular, no specific focusing is required inside the cryomodules.Depending on the transport line length, only a few quadrupole triplets are really necessary. Anintermediate analysis line could be useful in between the two cryomodules (at an energy of 5 MeV)in order to fully characterize the beam at the exit of the capture cavity.Another quite interesting feature is the beam shaping on the target. Depending on the target size andshape, the beam spot can be moved using a pair of deviating magnets. This easily allows coveringany area shape on the target. For example, an annular beam can be formed using only 0.05 Tdeviating magnets in the (xy) plane 90° out of phase and located 2.5 m away from the target. Thecorresponding beam envelope is calculated from the magnets to the target point(see below).

Beam envelopes using magnets to produce an annular beam on the target.

Cryogenic PlantThe required cryogenic power is of the order of 150 W at the working temperature of 1.9 Kincluding the static losses, the power coupler losses and a standard margin. The Hélial 2000liquefier from Air Liquide can fulfill these requirements. Using nitrogen pre-cooling, itsperformance can reach 130 l/h in the liquefying mode and over 400 W (@ 4.5 K) in therefrigeration mode. This liquefier has many interesting characteristics as the static gas bearingexpansion turbines and the industrial oil lubricated screw compressor. It offers the remotemonitoring from a distant location (control room) through a fully automated controller.Although the liquefier is the major component of the cryogenic plant, the overall plant shouldinclude the liquefier (150 W @ 1.9 K), a large dewar, a helium compressor, a storage tank forhelium gas, a pumping station, transfer lines for helium and nitrogen, a liquid nitrogen reservoir andall ancillary components : controller, gas purification, heater, etc…

ShieldingThe electron accelerator has to be shielded from radiological gamma rays that may be producedupon beam losses especially in the beam dumps. This is the only radiological hazard, since noneutrons or protons are emitted, only gammas by Bremsstrahlung. A first set of calculation has beencarried out to have a rough idea of what shielding would be required in order to achieve less than0.5 µG/h (Note that this level corresponds to what will be required in the future for a non-controlled(open) area ) outside the accelerator area. The shielding will be primarily due to the high-energy

beam dump losses (which are not yet defined). So the calculation assumes two variables : the lossesin the beam dump (labeled i1) and the losses in the deviating magnet (i2). In any case, these lossesare lower than the full beam current (500 µA).The figures below show the concrete thicknessrequired in the front direction (called the primary wall) and perpendicular to the beam axis(secondary wall) using a regular concrete (of density 2.35 g/cm3). In practice, a high-densityconcrete is generally used. In any case, this analysis shows that even in the extreme cases (wherethe shielding is designed assuming a full beam loss, which will never be the case), the primary wallwould not exceed 3 m of high density concrete (or 5 m of standard) and the side walls less than 2m. These figures will be reduced if one considers actual beam losses. In the same manner, theshielding has been evaluated along the transport beam line to the target area.

Thickness of the primary wall (regular concrete) as a function of beam current losses in the beamdump (i1) and in the bending magnet (i2) (upper). Same for the side wall (below).

Control-CommandThe proposed control-command for the driver accelerator can be done following a typicalarchitecture used for large machines. A possible solution would be based on UNIX stations (SunUltra 5), VME based CPU relying on the EPICS software and VME interface cards. This solution isbased on reliable technologies which are already implemented on many large equipment worldwide(DESY, LANL, etc…). The cryogenic system and the safety handling system would use each aspecific controller, but some basic information will have to be shared with the control-command.

Diagnostics & SafetyMany diagnostics are needed in order to continuously keep the beam in control. The vacuum levelat different locations should be permanently surveyed. Different other diagnostics like currentmonitors, gamma ray detectors, beam profilers, CCD cameras (light emission) and cryogenicdiagnostics are also required.For safety management, the beam current loss would be of primary interest (as seen from the aboveradiation analysis). Other safety trips may include radiation detectors, arc detectors and abnormalvacuum level.

ScheduleA high level schedule is shown in below (after completion of the TDS phase). First beam isexpected within two years after the project construction kick-off but the final beam delivered to thetarget would be envisaged 3 years after project approval (assumed to be the starting point – January2003).

PLANNING GENERAL DRIVER SPIRAL II

Période janv-03 avr-03 juil-03 oct-03 janv-04 avr-04 juil-04 sept-04 déc-04 avr-05 juil-05 oct-05

CAVITES CAVITES REALISEES

COUPLEUR COUPLEURS REALISES

CRYOMODULE CRYOMODULES TERMINES

SOURCES DE PUISSANCE HYPERFREQUENCE OPERATIONNELLE

INFRASTRUCTURES

MONTAGE TESTS FAISCEAU

Schedule of the electron accelerator construction.The reference design of the electron driver accelerator for the SPIRAL II project is asuperconducting linear accelerator. This allows continuous (CW) beam operation and a veryhigh efficiency (∼∼∼∼ 100% RF to beam). It also enables to use low and cheap RF power sourcesand to make profit of superconducting radiofrequency (SCRF) cavities operating at highgradients. Moreover, it offers the possibility of easy upgrading.

A new option for a deuteron driver in the framework of the LINAG project.The Linag phase I

In the framework of the LINAG project [4], it is envisaged in a first phase to accelerate a deuteronbeam at an energy of 40 MeV in an linear accelerator. This option is extensively described in thereport mentioned before. Its cost will be studied in the last part of this document as well as the twoother options.

SPIRAL II : Post-acceleration

The post-acceleration of RIBs produced by a new driver, other than the GANIL heavy ioncyclotrons raises the following questions :

• what is the energy range that can be reached in the cyclotron CIME by neutron-richisotopes?

• if this range is too limited for the various fields of physics, is it possible :-to re-inject the beams issued from CIME into one of the SSCs ?-to inject the beam issued from the source into the entire series of GANIL cyclotrons,skipping CIME ?

• what is the best lay-out for the target + source cave in order to :- preserve the development of the present SPIRAL installation ?- feed a low energy beam line ?

Post-acceleration

Post-acceleration by CIMEIt is to be remembered that the exotic ions exiting leaving the N+ source are in small

number. Therefore, the operation point of the source must be chosen for the most probable chargestate with the highest intensity. In what follows, we have considered that the source is a standard 14GHz ECR and we have collected a series of experimental results ( figure below) giving the

Room temperature 14 GHz ECR4 sources : Most probable charge

y = 0,7092x0,7835

y = 1,2757x0,6851y = 0,7277x0,8016

0

5

10

15

20

25

30

0 10 20 30 40 50 60 70 80 90 100

Atomic number Z

Mos

t pro

babl

e ch

arge

GAZmivocµfourSputdirectePuissance (GAZ)Puissance (mivoc)Puissance (µfour)Puissance (Sput)Puissance (directe)

Oven

Rod

Sputtering

Gas

MIVOC

optimum charge state obtained by different methods, as a function of the atomic number of the ionspecies. It is then easy, at least for post-acceleration by a single cyclotron without stripping, to havea good image of the maximum attainable energy.

CIME was optimized for ion masses below 100. Around this value and higher, the following tableindicates for some examples of neutron-rich isotopes, the maximum energy for the average chargestate read on the previous figure .

Element Z A Q average W max (MeV/n)Ni 28 68 13 9.6Ni 28 78 13 7.3Kr 36 90 15 7.3Kr 36 94 15 6.7Sn 50 128 18 5.2Sn 50 132 18 4.9Xe 54 140 20 5.4Xe 54 144 20 5.0

For these projectiles bombarding for example C or Pb targets, the coulomb barrier rangesfrom 3 to 5 MeV/n. Therefore, physics in this vicinity is possible, but the available energyrange is quite limited above.

Acceleration by CIME and the GANIL SSC

Cyclotrons operating with identical RF frequenciesIn this situation, there exists a series of conditions of compatibility between RF harmonic modes,and magnetic and RF frequency ranges.Only one case of direct re-injection of a CIME beam into an SSC is possible: CIME operating inharmonic mode 5 and SSC2 on harmonic mode 4. As seen on the CIME working diagram (figurebelow), this considerably restricts the operating range of CIME to a narrow band of injectionenergies 1.7 < W < 3.4 MeV/n. This leads to CCS2 output energies ranging from 11 to 21 MeV/nand requires a solid stripper anyway.

Diagramme de fonctionnement de CIME

0,6

0,8

1

1,2

1,4

1,6

1 2 3 4 5 6 7 8F/H (MHz)

0.10 0.15 0.20 0.25

0.30

0.35

0.40

0.45

0.50

0.55

H=4H=5 H=3

H=2

B (T)

Limitation V=34 kV

Limitation V=10 kV

ε=

W min 1.7 W max 3,4

Zone de compatibilité CIME- CSS2

CSS2W Min = 10W Max = 21

Any other case would require both stripping and energy degradation of the beam extracted fromCIME (provided the mean charge state obtained after the degrader would fit the RF frequencyconditions) , thus leading to a degradation of the beam characteristics and the resulting undesirableion losses.

Cyclotrons operating with independent RF frequenciesThen, the beam bunches accelerated by CIME would have to be submitted to the followingtreatment :

• first, to be de-bunched over as close as possible to 360° by an RF cavity (C1) operating onthe CIME RF frequency,

• second, to pass through a second cavity (C2) in order to reduce the energy spread generatedby C1,

• third, to be re-bunched by a third cavity (C3) over ± 3° of the RF frequency chosen forSSC2.

As an illustration, in the case of 132Sn, C1 would have to provide a linear voltage of ± 1.7 MV witha risetime of 2.7 ns. Since it is doubtful that this were technically possible, a sine wave of at leasttwice this amplitude would be necessary in order to debunch with an efficiency of about 50 %. Thesecond cavity C2 would have comparable requirements, while the third one C3 would have todeliver much stronger a voltage than the present rebuncher R2 since operating on a quasi-continuous beam. Adding to this the fact that C1 to C3 must be variable frequency cavities makesthe job not foreseeable in a five-year period.As a conclusion, re-injecting the beam from CIME into SSC2 covers an energy range from 11to 21 MeV/n. If higher energies are needed, injection into the whole series of the GANILcyclotrons is to be considered

Acceleration by the C0, SSC1 and SSC2In order to reach energies higher than 21 MeV/n, the atoms produced in the fissile target

would have to be ionized, then directed to the Co injection line (either C01 or C02) through adedicated beam line and accelerated through SSC1 and SSC2 with the stripper in between. Themaximum energies attainable would then be the classical ones, as indicated on the figure below

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100

Atomic number Z

W (M

eV/n

)

SSC1 and SSC2 maximum energy

for stable ion beams

No basic calculations are needed in this case, only two choices have to be made for the transportedbeams :

• charge state : the roughly 150 m long beam path would be less expensive as regards thevacuum requirements (charge exchanges on the residual gas) if the beam were issued as 1+ions from the source.

• energy : as will be seen in the next paragraph, the ions would have to run alongside theSSCs. The higher their magnetic rigidity, the less sensitive they would be to the SSCsleakage fields, therefore pleading in favor of a high extraction voltage from the ion sourceand (again) the lowest possible charge state.

Taking into account the fact that a new CIME central region can be designed for 65 kV, it seemssensible to direct the 1+ ions toward the C01 high voltage platform. This would also make the beamline shorter and simpler as opposed to an injection into C02

Conclusion and final remark

Post acceleration by CIME of the radioactive beams generated through fission ofuranium targets is limited to energies below 10 MeV/n. According to the ion species and to thepossibility to go to higher charge states at the expense of the intensity, there might exist anenergy gap between the CIME maximum energy and the SSC2 minimum one.

Energies above 21 MeV/n can be reached using the three GANIL cyclotrons as a post-accelerator. However, we want to stress the point that tuning and controlling severalcyclotrons in a row for very small intensity beams is not at hand. No strong experience isacquired yet, even on a single cyclotron like CIME and this procedure seems very un-appropriate.

The postacceleration using CIME and the combination of CIME + CSS2 is therefore limitedbelow 21 MeV/n as does not allow to cover the whole programme of physics described in thebeginning of the report.

The possibility of adding to the SPIRALII project a linear post accelerator up to energies of100 MeV/u is under discussion and consideration. This would allow to cover the wideprogramme of physics.

Implantation of SPIRAL II

The implantation of SPIRAL II has been devised by taking into account the main followingoptions :

• The radioactive ions will be produced inside a specific cave, located at the west side of theGANIL accelerator building. The advantages are numerous, keeping free the second cave ofthe SPIRAL facility as a spare of the cave used at present or for the development of newtechniques of ion production.

• The new SPIRAL II cave will be able to be built and commissioned without disturbing theSPIRAL functioning (otherwise, the adaptation of the SPIRAL cave 2 for SPIRAL II wouldhave stopped the use of SPIRAL for a long time)

• A new cave will be able to be designed according to the SPIRAL II constraints, whichwould not have been the case with the SPIRAL cave 2 (lack of room, existing walls,presence of SPIRAL equipments)

• Taking into account the other projects planned in the near future around SPIRAL as the lowenergy beam facility (LIRAT) or the equiping of cave 2.

• Allowing to simultaneaous use most of the facilities. For instance, using LIRAT with thebeam produced by the GANIL primary beam in the cave 1 and, simultaneously, providingexperiment area with the beam coming from the SPIRAL II cave after accelerating in CIME.

• Keeping free room for future extensions such as a booster for CIME or a post-accelerationin CSS2

• Disturbing the functioning of the present facility as little as possible. For instance, keepingfree the access to the accelerator hall for the trucks.

Technical options

Even if most of the technical options have no real influence on the implantation, a few must betaken into account .

• The principle of a two stage ionisation (mono-charge state source close to the productiontarget followed by a charge breeder) has been chosen .

• A mass separator with the highest possible efficiency is suitable to purify the beam used bythe low energy beam experiments and, also, to make easier the injection into CIME. Asecondary output would be useful to provide simultaneously the low energy experimentswith other type of ions. A low magnetic field dipole and a 30 kV source voltage could allowto reach an efficiency of 1000. The efficiency could be improved up to 2000 if the sourcevoltage was increased to 60 kV (that imposes a platform). The room for a very highefficiency separator will be kept.

• The ions must be identified before injecting them into CIME. The present identificationbench will be used, jointly with SPIRAL.

• The target will be vertically irradiated in order to simplify the safety problems.

Safety aspects

Up to now, few works have been carried out to evaluate the safety constraints and their influence onthe SPIRAL II layout. The only data available today concern the safety of the electron linearaccelerator (a 2m thick concrete wall around the accelerator seems sufficient) and the safety aroundthe targets for deuterons (neutrons flux, fission products, ...).One of the major problems not yet studied concerns the handling and the storage of the irradiateduranium targets. Our policy has been to keep open the different options, hoping that the room leftfree to install the equipments needed to respect the safety constraints would be sufficient. Thedetailed studies, carried out during the project design period, will allow to define these pointsprecisely.

Implantation Proposal

The schemes below show the implantation proposed for the electron option and for the deuteronoption (the linac tunnel is replaced by the cyclotron cave). The different parts of the facility are thefollowing :

1. Linear accelerator in a tunnel. The electron energy, of 45 MeV in a first step, can be increased by increasingthe cavity field. A 2 m thick roof covers the tunnel.

2. Linac gallery where the electronic devices (RF devices in particular) are installed.3. Transit gallery to access to the linac gallery and to the LIRAT experiment area from outside. A crane allows to

handle the equipments.4. Electron beam damper.5. Target and mono-charge state source at a bottom of a kind of well. The beam line is oriented vertically.6. Secondary well to handle and extract out the target after irradiation.7. Mass separator consisting in a low magnetic field dipole.8. Free room for very high efficiency mass spectrometer.9. Charge breeder.10. Separator secondary output to provide the low energy experiments (LIRAT) with species not used by CIME.11. Junction with the CIME injection line. The identification bench is adapted for use by SPIRAL I and II.12. Existing SPIRAL cave 113. SPIRAL cave 2 not yet equipped. Probably, this cave will be equipped with a two stage ionisation system.14. Low energy experiment area (LIRAT)15. Free room to install a booster to post-accelerate the beam coming from CIME or directly from the target via a

charge breeder.16. Beam line to possibly transport ions from SPIRAL target to GANIL injector or from CIME to CSS2 in order to

re-accelerate them.17. Limit of an external building for services which could be an extension of the present SPIRAL building.


The cost of SPIRAL II has been devised by taking into account the main following options:• The cost of the buildings has been deduced from the actual cost of the SPIRAL I building

(about 2 k€/m2). This building included an underground floor. For the cyclotron option, anunderground floor is needed. For the linac options, the cost should be roughly the same withor without underground floor (the cost of the excavation is compensated by lessradioprotection concrete). The basic networks and equipemnts (cranes...) are included.

• The cost of the target-source set and its infrastructures (caves, equipments, target handlingsystem, target storage, ...), estimated at 3.2 M€, has been extrapolated from the cost of theSPIRAL I one (2 M€), taking into account that the target-source set of SPIRAL II will belarger, with more safety constraints. However, the cost of a target retreatment facility hasbeen considered as an extra-budget (between 2 and 4 M€ according to the level ofradioprotection regulation), the basic option being to store the targets without retreatingthem.

• The cost of the driver has been devised as follows :! Cyclotron for 80 MeV deuterons: Its price (12 M€) has been got from the european

industry. It is valid for a turn-on-key cyclotron. This price could be reduced if somecomponents are taken in charge by GANIL (deuteron source, control system forinstance). A participation to the installation could reduce the price also.

! Linear accelerator for 45 MeV electrons: The design and the construction will be carriedon by French laboratories. The salaries of people working in these laboratories are nottaken into account in the cost. Consequently, its price (6.1 M€) includes mainly thesupplying of the components. Morever, the re-use of a part of MACSE (prototype of asuperconducting linear accelerator built in Saclay) could reduce the cost (between 1 to 2M€ could be saved).

! Linear accelerator for 40 MeV deuterons: As for the electron linac, the design and theconstruction will be carried on by European laboratories, that means the salaries ofpeople working in these laboratories are not taken into account in the cost and the priceincludes mostly the supplying of the components (18 M€). Moreover, the RFQ, pre-accelerating the beam going out from the source, is still in a conceptual design state. So,its price is not yet precisely estimated.

• The cost of the beam lines depends on their performances:! Beam lines from the driver to the target and from the target to CIME. These beam lines

are as sophisticated as the SPIRAL I ones. So; the same price as for SPIRAL I will beretained for these lines (75 k€/m);

! The 1/1000 separator has been estimated at 300 k€;! The price of the charge breeder is based on the market price of ECR sources (400 k€)! The beam line from the second separator output to the LIRAT line junction. This line is

just a transport line. Its price is lower (40 k€/m);! The beam line from CIME to CSS2 is partly a sophisticated line and partly just a

transport line. Its price reflects this fact (50 k€/m). The price of the corridor-likebuilding to cover the line with radiologic protection must be added (1 k€/m2);

• The cost of the control-system is relatively low (300 k€) because the drivers are simple tocontrol (constant energy, unique particule).

• The cost of the radioprotection is difficult to estimate. Indeed, up to now, there has been ereno real study of the constraints induced by the regulation, in particular concerning the use ofuranium in the target. The price which is proposed (1.5 M€) is just an extrapolation of thecost of the radioprotection for SPIRAL I (1 M€), without taking into account furtherconstraints like a confining vessel around the target.

• Few further expenses have been taken into account like travels (0,15 M€), site roads, carparks and green areas (0,15 M€).


Estimation of the costIn the next page, is given the estimated cost in millions of Euros (M €)

SPIRAL II BUDGET

(M€) 40 MeV deuteronlinac

45 MeV electronlinac

80 MeV deuteroncyclotron

Building / Infrastructure 5,6 3,2 4,3Driver 18,6 6,2 12,2Targets / Sources 1+ 3,2 3,2 3,2Source N+ 0,4 0,4 0,4Beam lines 4,6 4,3 4,3Radioprotection 1,5 1,2 1,5Control system 0,3 0,3 0,3Miscellaneous 0,3 0,3 0,3Hazards (10%) 3,5 1,9 2,6

TOTAL (M€) 38,0 21,0 29,1

MACSE re-use -2,3

GRAND TOTAL (M€) 38,0 18,7 29,1


In the following, one will find the list of the people involved in the Preliminary Design Study ofSPIRAL II.

International Advisory CommitteeG. de Angelis (INFN Legnaro), S. Brandenburg (KVI Gröningen), Ph. Dessagne (IReS Strasbourg),J.P. Gatesoupe (DSNQ/MSN CEA Saclay), W. Gelletly (University of Surrey), M. Huyse (KULeuven), B. Jonson (Göteborg), J. Martino (SPhN/DAPNIA Sacla), W. Mittig (GANIL Caen), Yu.Ts. Oganessian (JINR Dubna), M. Schädel (GSI Darmstadt), R. Julin (Jyväskylä)

Physics CaseF. Auger (Saclay), J.F. Berger (Bruyères le Châtel), B. Blank (CENBG Bordeaux), M.J.G. Borge(Madrid), A. Bracco (INFN), M. Chartier (CENBG Bordeaux), P. Roussel-Chomaz (GANIL), P.Dessagne (IreS Strasbourg), M. Girod (Bruyères le Châtel), S. Grévy (LPC Caen), F. Gulminelli(LPC Caen), M. Hellström (GSI Darmstadt), W.Korten (Saclay), D. Lacroix (LPC Caen), D.Lunney (CSNSM Orsay), M. Marques (LPC Caen), G. Neyens (KU Leuven), J.A. Pinston (ISNGrenoble), M. G. Porquet (CSNSM Orsay), N. Redon (IPN Lyon), P.H. Regan (University ofSurrey), O. Sorlin (IPN Orsay), C. Stodel (GANIL), C. Volpe (IPN Orsay), J.P. Wieleczko(GANIL)

Target-Ion source ensemblesN. Chauvin (CSNSM Orsay), S. Essabaa (IPN Orsay), F. Ibrahim (IPN Orsay), P. Jardin (GANIL),U. Koester ISOLDE), D. Ridikas (Saclay) H. Safa (Saclay), T. Lamy (ISN Grenoble), M G SaintLaurent(GANIL), ACC. Villari (GANIL)

Deuteron driverF. Loyer (GANIL), S. Brandenburg (Gröningen)

Electron driverJ.L. Biarrotte (IPN Orsay), P. Blache (IPN Orsay), L. Bourgois (Saclay), C. Commeaux (IPNOrsay), G. Devanz (Saclay), J.F. Gournay (Saclay), M. Jablonka (Saclay), T. Junquera (IPN Orsay),M. Lamendin (Saclay), M. Luong (Saclay), N. Pichoff (Saclay), M.Poitevin (Saclay), H. Safa(Saclay), H. Saugnac (IPN Orsay), C.Travier (Saclay)

Postacceleration of radioactive beamsE. Baron (GANIL), F. Chautard (GANIL), D. Jacquot (GANIL, F. Varenne (GANIL)

References

[1] Spiral II European RTT contract number ERBFMGECT980100, M.G. Saint Laurent, G.Lhersonneau, J. Aystö, S. Brandenburg, A.C. Mueller, J. Vervier[2] C. Lau, PhD, University Paris 7, IPN Orsay T 00 08 (2000) N. Pauwels, PhD, University Paris Sud, IPN Orsay T 00 12 (2000) F. Clapier et al., Phys. Rev.ST Accelerator and beams, 1 (1998) 013501 F. Ibrahim et al. Submitted to Euro. Phys. Journal B. Roussiére et al. Submitted to NIM[3] Module Accélérateur de Cavités Supraconductrices à Electrons, Internal Report DAPNIA/SEA92-09, 1992[4] High intensity beams at GANIL and future opportunities : LINAG, G. Auger, W. Mittig, M.H.Moscatello, A.C.C. Villari Ganil Report R01 02


STATUS OF EXPERIMENT

Title: ALICE

Date of the first CSTS presentation: November 18th 1997

Experiment carried out at: CERN

Spokes person(s): J. Schukraft

Contact person at SPhN: A. Baldisseri

Experimental team at SPhN: A. Baldisseri, H. Borel, E. Dumonteil (PHD student), J. Gosset, F. Staley

List of DAPNIA divisions and number of people involved: SEDI (7), SIS (3), SACM (1). All the people

are not involved full time.

List of the laboratories and/or universities in the collaboration and number of people involved: In France

IPN Lyon, IPN Orsay, Clermont−Ferrand, SUBATECH Nantes, Strasbourg, for a total of ~73 institutes

over ~27 countries and ~1000 people involved.

SCHEDULE

Starting date of the experiment [including preparation]: April 2007

Total beam time allocated: ~1 month/year of heavy ions

Total beam time used: Several years

Data analysis duration: One year for 1 month data taking

Final results foreseen for: 2012 (?)

BUDGET Total Already UsedTotal investment costs for the collaboration: 915 kEuro (6 MF) 305 kEuro (2 MF)Share of the total investment costs for SPhN: ~1% (~7% of muon arm)Total travel budget for SPhN: ~46 kEuro (300 kF)/year

Please include in the report references to any published document on the present experiment.

Among the large number of reports written by the ALICE collaboration concerning all the detectors,the followings are the most important ones for the Dimuon Arm :

The Forward Muon Spectrometer, CERN/LHCC 96−32Dimuon Arm Technical design report, CERN/LHCC 99−22Addendum to the Dimuon Arm Technical Design Report, CERN/LHCC 2000−046



Title: PHENIX

Date of the first CSTS presentation: December 5th 2000

Experiment carried out at: Brookhaven (USA)

Spokes person(s): W. Zajc

Contact person at SPhN: A. Baldisseri

Experimental team at SPhN: A. Baldisseri, H. Borel, Y. Cobigo (PHD student), J. Gosset, F. Staley

List of DAPNIA divisions and number of people involved: No technical commitment

List of the laboratories and/or universities in the collaboration and number of people involved: In

France IPN Orsay, SUBATECH Nantes, Clermont−Ferrand, Polytechnique for a total of 52 institutes

over 11 countries and ~430 people involved

SCHEDULE

Starting date of the experiment [including preparation]: First run in June 2000

Total beam time allocated: ~8 months per year

Total beam time used: ~6 months for heavy ions

Data analysis duration: Several years

Final results foreseen for: 2007 (?)

BUDGET Total Already UsedTotal investment costs for the collaboration: 76 kEuro (500 kF) 76 k EuroShare of the total investment costs for SPhN: 25% of the french contribTotal travel budget for SPhN: ~ 30 kEuro (200 kF)


The publications concerning the RUN 1 (year 2000 data taking) analysis are : 6 accepted in PRL,4 submitted (3 PRL and 1 PRC), 3 papers in progress. For the RUN 2 the analysis work is under way.

Status Report on ALICE and PHENIX experiments

CSTS du SPhN June 2-3 2002

The general physics topic of our group is the search and study of the Quark Gluon Plasma (QGP). Weare mainly interested in the signatures coming from the resonances (J/ , ) that are studied in the dimuonchannel. We are involved in two programs : ALICE (LHC at CERN) and PHENIX (RHIC at Brookhaven).We briey present a status report on both activities.

ALICE

The SPhN group is involved in the design and construction of the large tracking chambers of the Dimuon Arm.The 10 tracking chambers of the Dimuon Arm consist of 5 stations (2 chambers each) of cathode pad chambers.The design of the 3 largest stations is modular : each chamber is made of several modulus (called slats) ofseveral dimensions. The dimensions of the slats are : 40 cm high and 80 to 240 cm long (active area), in orderto cover the spectrometer acceptance (2Æ < < 9Æ). Four laboratories are in charge of the production of thoseslats : INFN Cagliari, SUBATECH Nantes, PNPI Gatchina and DAPNIA/SPhN Saclay. The IPN Orsay isin charge of the electronics of all tracking chambers.

Saclay is responsible for the support of the slats and will construct about 40 slats (over 160). We are alsoin charge of the cooling and the integration of the chambers in the Dimuon Arm, and for that we are in stronginteraction with the CERN technical team.

The main characteristics of a slat (resolution, eciency, gain, ...) have been validated in October 2001 testbeam at CERN with a 240 cm long prototype (g. 1). The Saclay group had a leading position in the testbeam data analysis, since we developed most of the software. The resolution obtained is ' 70 m and 97%eciency over a large HV plateau. These results fulll the requirements of the Dimuon Spectrometer.

Some other prototypes have also been used to validate the building process, the gluing, the mechanicalproperties, etc. The nal R&D and validation of all the technical solutions will be nish this year.

The start of the production is foreseen in the four laboratories in the beginning of 2003. Each laboratoryhas to be equipped with an assembly hall which fulll all the requirements (temperature and humidity control,overpressure) with all the tooling needed. For Saclay, a location is foreseen and a price estimation is under way.Some part of the tooling (granite table) is already ordered.

The slats supports are one of the most important items under the Saclay responsibility (design and con-struction). The retained solution is to use a honeycomb sandwich with carbon ber skins. The use of carbonber ensures a good stiness and a low thermal dilatation. The overall dimensions for the biggest one is 6m high and 3 m wide with a thickness of 15 mm and with a planarity within 10 mm. It must be also rathertransparent to muons (< 0.3 % radiation length). Each support hangs the slats (250 Kg with cables) for a halfchamber. The survey market is already done and the call for tender is expected for June 2002. Neverthelessthe price estimation from dierent companies seems to be signicantly higher than expected.

An air cooling solution has been adopted after a large simulation work done at Saclay. The integration ofthis solution in the experiment setup is under way. It is particularly delicate for the Station 3 which is insidethe dipole magnet due to a lack of place.

Concerning the software, we are involved in the simulations, in particular in the tracking reconstruction andrecently also in the implementation of a realistic chamber response.

Presently 3 physicist (among 4) and one student are mainly involved in ALICE. One of us (F. Staley) is theproject leader of the Dimuon Arm.

1

Figure 1: 240 cm long slat during the October 2001 test beam at CERN T10. We can see the electronicsconnectors on top and bottom of the slat.

PHENIX

The RHIC heavy ion collider runs at intermediate energy between the SPS and the LHC, this provides an uniqueopportunity to explore the properties of the QGP. Among the four experiments dedicated to this physics atBrookhaven, PHENIX is the only one devoted to the resonances study in the dimuon channel using two dimuonarms.

The RHIC accelerator started on June 2000 at an energy ofps = 130 GeV/nucleon-nucleon with a low

luminosity and without the muon arms. These data have been quickly analyzed and a lot of results on theglobal variables behavior has already been presented in several conferences. On one hand, some of the resultsare on the continuity of the SPS ones : energy density, temperature, etc. On the other hand, the most striking issurely the p

Tdependence of the particle production which shows a depletion at high p

T(g. 2). This behavior

suggests a jet quenching with an energy loss of 0.25 GeV/fm. For the second run started on summer 2001the nominal energy (

ps = 200 GeV/nucleon-nucleon) has been reached with a more complete setup with the

South Muon Arm. Since it was the rst run with a Dimuon Arm, most of the eort has been put on the codedebugging, chambers alignment, etc, in order to extract the resonances signal. This work is still under way. Forthe next run beginning in fall 2002, the two muons arms will be operational, increasing considerably the physicspotential, in particular concerning the J/ , where the theoretical predictions are controversial (suppression orenhancement). We expect to put serious constraints on the dierent theoretical models.

The French participation to PHENIX is done in the framework of PHENIX-France formed by several lab-oratories of the IN2P3 (IPN Orsay, Clermont-Ferrand, SUBATECH Nantes, Ecole Polytechnique) and theDAPNIA/SPhN. PHENIX-France contributes to the electronics of the tracking chambers of the North DimuonArm. The technical commitment of the electronics construction and test is under the responsibility of the LLRlaboratory of the Ecole Polytechnique. The SPhN participation is limited to a nancial contribution of 25% ofthe total cost of the electronics.

The installation and test of the North Dimuon Arm electronics is now under way at Brookhaven and it will benished well before the beginning of the next run. Our group participates actively to this task in collaborationwith the other laboratories of PHENIX-France.

The main contribution of the SPhN to PHENIX is focused on the data analysis. One important contributionwas done in the track tting using the Kalman lter, where a C++ code has been setup and tested. Theimplementation of this code in the PHENIX software is nearly nished.

Presently 1 physicist (among 4) and one student are fully involved in PHENIX.

2

Figure 2: Transverse momentum spectra for the 0 in the 10% most central events

3



Title: Study of exotic nuclei at Ganil

Date of the first CSTS presentation: June 6th, 2000

Experiment carried out at: Ganil

Spokes person(s): A. Gillibert, L. Nalpas

Contact person at SPhN: L. Nalpas

Experimental team at SPhN: N. Alamanos, F. Auger, A. Drouard, A. Gillibert, V. Lapoux, L. Nalpas,

E.C. Pollacco, R. Raabe, F. Skaza, J.-L. Sida.

List of DAPNIA divisions and number of people involved: SEDI (2)


IPN Orsay (3), Ganil (3), Sao Paulo Univ. (1), Ioaninna Univ. (1)

SCHEDULE for E417 experiment

Starting date of the experiment [including preparation]: September 2002

Total beam time allocated: 24 UT

Total beam time used:

Data analysis duration: 24 months

Final results foreseen for:

BUDGET Total Already Used

Total investment costs for the collaboration:

Share of the total investment costs for SPhN: 14 kEURTotal travel budget for SPhN: 6 kEUR


Study of exotic nuclei at Ganil

Abstract

The main objective of these studies is the understanding of the detailed structure of exotic nuclei far fromthe stability for which strong discrepancies still exist between the different theoretical predictions. Withinthis framework, the study of halo nuclei is the core of our experimental program. The exotic nuclei areproduced on line and interact with known light targets (1H, 12C). Elastic and inelastic measurements giveus information on the entrance channel, excited states, B(E2) values and densities. On the other hand,transfer reactions give us information on the wave functions and the spectroscopic factors. Over the lastfew years, we have investigated light exotic nuclei through Coulomb excitation measurements – 30,32Mg -(p,p’) reactions - 6,8He, 10,11C, 20,22O – and transfer reactions - 6He(p,t)4He – leading to 2 PhD theses and 10publications. The F. Skaza’s PhD is in progress on the 8He(p,p’) data. Recently, we have obtained beamtime from the Ganil PAC to study the 13Be unbound nucleus through (d,3He) transfer reaction (see theincluding proposal). Over the next few years, we want to pursue this experimental program with Mustusing Ganil and especially Spiral beams taking advantage of the new Vamos facility. With Must II, wewill be able to improve the angular coverage and the coupling capabilities with other devices such asmass and gamma spectrometers.

Ganil proposal:Structure of 13Be with (d,3He) transfer reaction

1. Motivations

The structure of the unbound systems 10Li and 13Be plays a major role in the description of theBorromean nuclei 11Li and 14Be. These nuclei exhibit a peculiar structure with 2 loosely bound neutrons(core+n+n) while each binary subsystem (core+n, n+n) is unbound. As for 11Li, the 14Be nucleus is a goodcandidate for 2-neutron halo: its ground state is weakly bound (S2n ∼ 1.34 MeV) and its subsystem 13Be isunbound.

In the N=7 neutron rich nuclei, the lowering of the 2s1/2 single particle neutron state respectively tothe 1p1/2 state was proposed to explain the abnormal parity 1/2+ of the 11Be ground state. In this context,a s-wave resonance has been proposed for the 10Li ground state [Thom94]. For a similar situation in theN=9 region, the unbound 13Be ground state would be Jπ = 1/2- instead of 1/2+ [Labi99], as a consequenceof such a parity inversion. A same trend has been observed for the 1+ level at 1.28 MeV in 14B [Aoi97].

The inversion of 1p1/2 and 2s1/2 neutron shells in 13Be is predicted in many theoretical works[Descou95, Ren95, Thom96, Labi99] to explain the known properties of 14Be. Strong correlationsbetween the extra neutron and the core should be responsible for this phenomenon.

From an experimental point of view, some excited states have been observed so far, especially astate at 2.01 MeV over the threshold which was assumed to be the neutron ν1d5/2+ state [Ostro92, Kor95,Belo98]. At lower excitation energy, the situation is not clear. Due to a large background, the results forthe neutron pick-up 12Be (d,p) 13Be were inconclusive [Kor95]. A state was observed in the multi-nucleon11B(14C,12N)13Be transfer reaction [Belo98], which was assumed to be the ground state, unstable with 0.80MeV over the threshold and a width equal to 1 MeV. The tentative assignment J = 1/2 was done withoutdiscrimination between ν1p1/2 and ν2s1/2. Recently, the relative velocity spectrum of the 13Be decayfragments, 12Be and n, was measured at MSU[Thoen00], after the fragmentation of a 18O projectile atintermediate energy. The results suggest the existence of a low energy resonance, 200 keV over thethreshold. The best fit was obtained for a s resonance, however the data do not exclude a p state (fig.1 in[Thoen00]). Finally, the nature of the state still remains an open question. The level scheme proposed in

[Thoen00] is presented in figure 1. If the state at 0.8 MeV corresponds to the ν1p1/2 orbit, that levelscheme do not support the level inversion predicted for example in [Lab99].

Figure 1: Level scheme of 13Be with respect to 12Be+n as proposed in [Thoen00]

An alternative way to get spectroscopic information about 13Be would be the measurement of theproton pick-up 14B(d,3He)13Be transfer reaction around 40 A.MeV. In the standard shell model, the protonwill be picked in the π1p3/2 shell and the l = 1 transfer coupled to the 2- spin of 14B ground state will onlyselect positive parity states. In a one-neutron removal experiment (14B,13B) [Gui00], the neutron valencein the 14B ground state wave function was shown to be mainly in the ν2s1/2 (89%) and ν1d5/2 (11%)orbits.

That seems to be a very good opportunity to see a s resonance close to the threshold and to fix itsparameters.

However in case of the shell inversion, that s resonance will not be the ground state of 13Be. It will benecessary to study an other transfer reaction, the 1 neutron stripping 12Be (d,p) 13Be, much less selective.In that reaction, it might be a problem to separate the contributions of the s and p resonances close to thethreshold, broad and close to each other, if one of them is not already known. On fig.2 from [Thoen99], asimilar problem is shown for 10Li if the two contribution have to be separated. For that reason, the14B(d,3He)13Be pick-up may be seen as a necessary preliminary step to the 12Be (d,p) 13Be stripping.

Figure2 : Relative population of the s-wave (dot-dashed) and the p-wave (dashed) deduced for 10Li in [Thoen99].

12Be+n13Be

<0.2

2.0

0.8

2.1

2.72+

0+1/2+

1/2

5/2+

2. Experimental details

In this experiment, we plan to measure coincidences between 3He with MUST and 12Be from thedecay of 13Be with SPEG. The spectrometer will work in the achromatic mode. Due to the high Bρ values(around 2.7 T.m) and the need for an efficient beam-tracking device with two CATS detectors before thetarget, SPEG will be used rather than VAMOS. With MUST, the position and energy of light chargedparticles will be measured with a good accuracy, details can be found in [Blum99].

Figure 3 : Kinematics in the laboratory frame of 3He ejectiles in (d,3He) transfer at 42 MeV/nucleon for the 13Be ground state.The corresponding center-of-mass angles are indicated on the curve.

Figure 3 shows the energy and angular ranges expected for this reaction. Due to the inversekinematics, the reaction products are focused at forward angles. For the l = 1 transfer, the domain ofinterest goes up to 15 degrees in the center-of-mass frame. The measurement will be performed with onesetting of SPEG at 0 degrees and the standard ion identification. The different reactions will be separatedin the focal plane and compared to the beam momentum reference po, with δ = (po-p)/po = +2.9% for(d,p), -3.4% for (d,t) and +18.4% for (d,3He). In the latter case, the neutron emission will scatter the 12Beions between +6% and +11% as shown on fig.4, still within the relative momentum acceptance of SPEG.However, the acceptance is not large enough to measure the 3 reactions simultaneously and we shallconcentrate on the (d,3He) transfer.

The low energy A=3 particles (3He and 3H ) will be identified by energy and time-of-flightmeasurements in the 300 µm thick silicon layer of MUST. The start of the timing will be given by one ofthe two beam detectors, called CATS, usually used to reconstruct the position and the incident angle ofthe projectile on the target. 3H will be rejected by the coincidence and identification of 12Be with SPEG.

We will use a very thin CD2 foil of 0.26 mg/cm2 built at Orsay, in order to minimise the angularand energy straggling inside the target. With MUST and that same target, an energy resolution better than400 keV was obtained for the reconstruction of 10Li in the previous 11Be(d,3He)10Li experiment [Pita00].It is enough to separate the two expected s and d resonances on a spectrum similar to fig.2, but with a dresonance at higher energy (2 MeV over the threshold).

14B + d 42 A.MeV

( d , 3He )

ΘHe (degres lab)

EH

e (M

eV)

2.8 5.7 7.110. 11.4

14.3

17.2

21.7

6

8

10

12

14

16

18

0 5 10 15 20 25 30 35 40

3. Beam time request

The 14B beam will be produced by fragmentation of the (63 MeV/nucleon) 18O primary beam. With anintensity corresponding to 400 W on a 513 mg/cm2 thick carbon target (SISSI), the LISE code predictsabout 2 105 14B ions per second at 42 MeV/nucleon. The secondary beam has to be purified with a 1.5 mmthick aluminium degrader. With these conditions, the main contaminants will be the 16C ion at about thesame rate and the 12Be ion at a factor of 10 below than the 14B rate. The total counting rate does notexceed 5 105 pps.

The beam time request is based on a mean transfer cross section around 1 mb/sr taken from the11Be(d,3He)10Li experiment [Pita00], an intensity of 2 105 pps and a 0.26 mg/cm2 thick CD2 target. Withthe SPEG spectrometer at zero degrees, the 12Be fragments will be detected with a high efficiency close to1. With these conditions, we expect about 50 coincidences per day. With a more detailed simulation andthe angular distribution cut in 2 °cm wide slices, we obtain on fig.5 the expected cross section. Thisshould be accurate enough to check that we deal with a l = 1 proton pick-up.

The total integrated statistics for 21UT is 306 coincidences. That is the total statistics for theexcitation energy spectrum we want to analyse.

Figure 4 : Be ejectiles at 42 MeV/nucleon with the momentum variation in abscissa (p-po)/po , po being the momentum of thebeam; (up) (d,α) transfer and (d,3He) transfer assuming 13Be is bound, the dots correspond to the other one on fig.2; (down)12Be after neutron emission, we see at 4 angles the extreme values for δp/p corresponding to a 13Be emitted at the same anglefor a given excitation energy. The total momentum acceptance of the SPEG spectrometer is ± 3.5%. The opening angle due tothe neutron emission is 1 degree.

14B + d 42 A.MeV

δp/p (%)

θ Be

(deg

res

lab)

( d , α ) ( d , 3He )

δp/p (%)

θ 12B

e (d

egre

s la

b)

E* = 0. MeV

0

0.5

1

1.5

2

2.5

3

14 15 16 17 18 19 20

0

0.5

1

1.5

2

2.5

3

5 6 7 8 9 10 11 12

Figure 5 : Angular distribution for a l = 1 transfer in 14B(d,3He)13Be; the distribution is cut in angular slices from 0 to15° cm with a width equal to 2° cm. The vertical error bars is the statistical error resulting from a simulation with theexperimental device and 7 day beam time.

Bibliography

[Aoi97] N. Aoi et al., Nucl. Phys. A616 (1997) 181c.[Belo98] A.V. Belozyorov et al., Nucl. Phys. A636 (1998) 419.[Blum99] Y. Blumenfeld et al., Nucl. Inst. Meth. A 421 (1999).[Descou95] P. Descouvemont, Phys. Rev C 52 (1995) 704.[Gui00] V. Guimaraes et al., Phys. Rev. C 61 (2000) 064609.[Kor95] A.A. Korsheninnikov et al., Phys. Lett. B 343 (1995) 53.[Labi99] M. Labiche et al., Phys. Rev. C 60 (1999) 027303.[Ostro92] A.N. Ostrowski et al., Z. Phys. A 343 (1992) 489.[Pita00] S. Pita, PhD thesis, University of Paris VI, 2000 (E319 Ganil experiment).[Ren95] Z. Ren et al., Phys. Lett. B351 (1995) 11.[Thom94] I.J. Thompson and M.V. Zhukov, Phys. Rev. C 49 (1994) 1904.[Thom96] I.J. Thompson and M.V. Zhukov, Phys. Rev. C 53 (1996) 708.[Thoen99] M. Thoennessen et al., Phys. Rev. C 59 (1999) 111.[Thoen00] M. Thoennessen et al., Phys. Rev. C 63 (2000) 014308.

14B ( d,3He ) 13Be

l = 1

θcm (deg)

dσ/d

Ω (

mb/

sr)

1

0 5 10 15


RESEARCH PROPOSAL

Title: Measurement of neutron flux at MEGAPIE

Experiment carried out at: PSI, Zurich, Switzerland

Spokes person(s): F. Marie

Contact person at SPhN: F. Marie

Experimental team at SPhN: M. Fadil, G. Fioni, S. Leray, A. Letourneau, D. Ridikas, H. Safa


SIS(3)


CEA/Cad DEN/DER/SPEX/LPE(2), PSI(2), SUBATECH/Nantes(3)

SCHEDULE

Possible starting date of the project and preparation time [months]: juin2002-dec2005

Total beam time requested:n/a

Expected data analysis duration [months]: 12

REQUESTED BUDGET

Total investment costs for the collaboration: n/a

Share of the total investment costs for SPhN: 120Keuro

Investment/year for SPhN: 60Keuro

Total travel budget for SPhN: 40Keuro

Travel budget/year for SPhN:5-10-5-15

If already evaluated by another Scientific Committee:

If approved Allocated beam time: Possible starting date:

If Conditionally Approved, Differed or Rejected please provide detailed information:

PROPOSITION AU CONSEIL SCIENTIFIQUE ET TECHNIQUEService de Physique Nucléaire, CEA/Saclay,3 et 4 juin 2002

Measurement of the neutron flux at MEGAPIE

M. Fadil, G. Fioni, S. Leray, A. Letourneau, F. Marie*, D. Ridikas, H. SafaCEA/Saclay, DSM/DAPNIA/SPhN, 91191 Gif-sur-Yvette, France

C. BlandinCEA/Cadarache, DEN/Cadarache/DER/SPEX/LPE, St Paul-les-Durance, France

F. Groeschel, E. LehmannPSI, Zurich, Switzerland

Abstract: We propose to measure the neutron flux in the Pb-Bi MEGAPIE target at PSI. Fissionmicro chambers developed in the frame of the Mini-Inca project at ILL will be placed in thecentral rod of the target to determine both thermal and fast components of the neutron flux. Inprinciple, both time- and space-dependent variations of the flux will be on line monitored with aprecision better than 10%.

Introduction

In the frame of the MEGAPIE (from MEGAwatt Pilot Experiment) project [1], a Pb-Biliquid target will be installed at the SINQ spallation neutron source [2] of the PaulScherrer Institut. It will assess experimentally the performance of a 1 MW liquid targetduring 4-9 months irradiation period. This will be an essential step for the furtherdevelopment of high power targets in the 20-50 MW range required for AcceleratorDriven Systems and for high intensity spallation neutron sources.Among all the uncertainties related to the operation of Pb-Bi spallation targets, two topicshave kept our interest :

• the formation of 210Po which generates a high radio-toxicity• the knowledge of the neutron flux, both in energy and in absolute intensity,

which will arise from the spallation target. This will influence the efficiency andperformance of Accelerator Driven Systems dedicated to nuclear wastetransmutation and also to intense neutron sources for material irradiation.

In order to evaluate the radiation hazard of Po, the Mini-Inca project has alreadypresented a proposal [3] to the CSTS to measure 209Bi neutron capture cross sections.Since 2001, several experiments are in progress at ILL and at IRMM/Geel to measure theneutron capture branching ratio of 209Bi to the ground and metastable states of 210Bi, atthermal, epi-thermal and fast neutron energies.

As long as the transmutation studies are concerned, the Mini-Inca group has developed,in collaboration with CEA/Cadarache, a new type of fission micro chambers [4], whichhave been tested during 26 days in December 2001 in the V4 channel of the High FluxReactor at the Institut Laue Langevin in Grenoble. The outstanding performancesobtained for our new fission chambers in a very intense neutron flux have validated a

* Corresponding author, [email protected]


new method to determine the transmutation potential of minor actinides. The intensity ofthe flux has been measured to be 1.8x1015 n/s/cm2 with a relative uncertainty of 4%. Weshould stress that it is the first time that a fission micro-chamber can be operated in sucha high neutron flux, with a gain of almost one order of magnitude on the operating fluxlevel.

The excellent results on the neutron flux measurements at ILL and the wish to extend ourmeasurements to a wider range of neutron energies, lead us naturally to propose asolution for measuring the neutron flux in the Pb-Bi liquid target at MEGAPIE. After theinitial proposal made in December 2000, we can now finalize this concept thanks to theprogress that we have recently accomplished.

Experimental Method

In Figure 1, the flux as a function of the neutron energy is plotted for different distancesfrom the beginning of the spallation target. The detailed simulations were performed byusing the MCNPX code obtained from LANL (USA), which actually combines MCNPand LAHET codes [5].

Figure 1: Volumetric neutron fluxes inside the Pb-Bi target at 5, 15, 25, 35, 45, 55,65cm (blue dashed curves from top to bottom ) with respect to the very bottomposition (beam interaction with window). Contribution of thermal neutrons may varyfrom ~10% to ~85% respectively. Solid red curve corresponds to the average neutronflux all over the Pb-Bi target (100cm long). See Figure 2 for a detailed geometrypresentation.


Due to the presence of a reflector of D2O around the SINQ target (see Figure 2), theresulting neutron spectrum inside the Pb-Bi target is composed of 32% of thermalneutrons.

Figure 2: 3D geometry of Pb-Bi target (violet) surrounded by heavy water tank(green) as entered for MCNPX simulations. A white circle in the center indicates thelowest previewed position at, say, 35cm, where the micro-chambers could be placed.

As a matter of fact, the thermal flux of MEGAPIE is of the same order of magnitude asthe flux measured at the upper position of the V4 beam tube at ILL, i.e. of the order of~1014 n/s/cm2 and most of the technical solutions developed to measure the neutron fluxat ILL will apply to the experimental conditions of MEGAPIE.

We propose to implement our fission micro chambers as neutron detectors in order tomeasure on line the flux in the MEGAPIE target. Due to the small place available in thecentral rod of the target, i.e. ≈1 cm internal diameter (see Figure 3), small detectors of thesame type as we operated at ILL (U-235 deposit, 4 mm diameter and 2 cm length) arecertainly good candidates. To our knowledge, in the present configuration, there is noother alternative way to obtain needed information on the neutron flux on line.

Monte Carlo simulations with MCNPX show that at a distance of 35 cm from the Pb-Bitarget entrance in the central rod the neutron flux is ~1.7x1014 n/s/cm2 with nearly 42%thermal. We have also done some estimations of the expected saturation current fordifferent fissile isotopes (see Table 1)

The expected saturation current in these fission chambers should be of the same order ofmagnitude as for the ILL irradiation ( ≈40 µA). As described in Figure 3, these chambersdisposed at different distances from the target entrance in the central rod will allow tomeasure the spatial variations of the thermal flux, tacking benefit that the calibration of


these chambers is already determined at ILL. More over, due to the much smaller burn upof U-235 compared to ILL conditions during MEGAPIE operation (less than 0.3%/day),these chambers will provide also an on line monitoring of the time dependency of thetotal neutron flux. We estimate that during 9 months of operation nearly 35% of initial U-235 mass still will be left as presented in Figure 4. This burn-up rate is low enough tomeasure on line flux variations due to, for example, the beam intensity-energy variationsor beam trips. At the same time, this burn-up rate is high enough to re-calibrate ourdetectors if necessary as it was done during the irradiation at ILL.

20 cm

Proton beam590 MeV1.8 mA

4 m

20 c

m

Fission chambers

Pb-Bi liquid target

2 m

D2O tank

Central rod

Target shieldingand heating

1cm

Figure 3: Scheme of a possible positioning of the neutron detectors in the MEGAPIEtarget .

No shielding With 2 mm Gd shielding

<σ> [b] <σ>.φ Is [µA] <σ>[b] <σ>.φ Is [µA]235U (ILL) 500 3.2 1016 44235U (MEGAPIE) 270 2.7 1016 37 9.4 3.4 1014 0.5232Th // 12. 10-3 1.2 1012 34.10-3 1.2 1012

238U // 40. 10-3 4.0 1012 0.1 3.6 1012

239Pu // 404 4.0 1016 10.9 4.0 1014

Table 1: Expected saturation current in the fission micro chambers during irradiation in MEGAPIE fordifferent fissile deposits (m=20 µg) compared to the current measured during ILL irradiation, with orwithout thermal neutron shielding.


Figure 4: Burnup of U-235 as a function of irradiation time inside Pb-Bi target at35cm.

Figure 5: Flux at 35cm without and with natural Gd envelope (2mm thick).

For the measurement of the epi-thermal component of the neutron flux, we propose toinstall, at a low position in the central rod, a fission chamber with 200 µg of U-235


shielded with a 2 mm thick Gd cover. The small value of the measured current ( ≈5 µA)will require a more precise acquisition electronics ( Aµ01.0± ). Figure 5 presentsexplicitly the calculated neutron flux seen by our micro-chambers without and with Gdenvelope. We note separately, that 2mm thick natural Gd coating will be sufficient toattenuate thermal neutrons at least up to the level of 10-5 for nearly 9 months.Finally, chambers without deposit will be placed together with “active” chambers inorder to measure the background generated by photoelectric effects from gammainteracting with the gas or structural materials of the fission chambers.

We believe that in this way both thermal and fast components of the neutron flux couldbe determined experimentally including their time- and space- variations on line.

ConclusionsThe neutron flux measurements we propose to perform in the MEGAPIE target are of agreat interest for the knowledge of the realistic irradiation conditions in the future hybridsystems dedicated for the transmutation of nuclear waste as well as for intense neutronsources for irradiation based on spallation reactions. These measurements will bring amuch better understanding of the microscopic processes involved in spallation targetsand will provide highly requested quantitative data to test simulation codes, both forneutron generation and transport in realistic geometry. These measurements will alsooffer a unique and promising possibility to monitor on line variations of the neutron fluxduring entire MEGAPIE operation.

Budget, schedule and manpower

8 neutron detectors with cables 45 KeuroElectronic and acquisition 70 KeuroGadolinium filter 5 KeuroMechanics + flux monitors 10 Keuro

130 Keuro

Table 2: Estimation of the investment budget for the implementation of the neutron detectors

From the technical point of view, the DAPNIA will have the responsibility of theconstruction of the fission micro chambers (in collaborartion with CEA/Cadarache) andthe flux measurements and analysis at PSI. The implantation of the detectors within thecentral rod of the target will be performed in collaboration with SUBATECH/Nantes.All the detectors should be installed by may 2003. The operation of the beam on theMEGAPIE target should start in march 2005 for 9 months. The manpower for ourparticipation in the flux measurement for MEGAPIE, including developpements of themicro fission chambers, is a PostDoc or a physicist in 2003 and a PhD student.


References [1] M. Salvatores, G.S. Bauer and G. Heusener, The MEGAPIE Initiative , Report MPO-1-GB-6/0_GB, Paul Scherrer Institute, Zurich (1999).[2] G.S. Bauer, The Swiss Spallation Neutron Source SINQ, Proc. Of ICONE 8,Baltimore (2000).[3] G. Fioni et al., Measurement of 209Bi thermal neutron capture branching ratio,Proposition au CSTS du DAPNIA/SPhN ,5/12/2000.[4] M. Fadil et al., Development of fission micro-chambers for nuclear waste incinerationstudies, Nucl. Instr. & Meth. A 476 (2002) 313; M. Fadil, PhD thesis in preparation.[5] D. Ridikas et al., in preparation for a publication.


LETTER OF INTENT

Title: Measurements in the future TRADE project

(TRiga Accelerator Driven Experiment)

Experiment carried out at: ENEA Casaccia-Roma (Italy)

Spokes person(s): C. Rubbia ?

Contact person at SPhN: S. Andriamonje

Experimental team at SPhN: S. Andriamonje, Y. Giomataris (SEDI), J. Pancin

List of DAPNIA divisions and number of people involved: SPhN (2), SEDI (3)

List of the laboratories and/or universities in the collaboration and number of people

involved: (possible)

ANSALDO (2), CEA-DEN-Cadarache (6), ENEA-Casaccia-Roma (18),

CERN-Geneva (5)

SCHEDULEEstimated total duration of the proposed experiment: 6 years

Possible starting date of the experiment:Spring 2003

Expected duration of the data analysis: 6 years

ESTIMATED BUDGETTotal investment costs for the collaboration:

Share of the total investment cost for SPhN:

Total Travel Budget for SPhN: 15 k /year

Measurements on the future TRADE project (TRiga Accelerator Driven Experiment)

Abstract:

The TRADE [1] project, which is part of the European Roadmap [2] towards thedevelopment of Accelerator Driven Systems (ADS), foresees the coupling of a 110 MeV,2 mA proton cyclotron with the core of a 1 MW TRIGA research reactor.The aim of this letter of intent concerns the experimental measurement andinstrumentation programme for the proposed project. That concerns in particular, thedevelopment of different techniques such as a micro fission chamber, a gamma rayspectroscopy associated with “fast rabbit” system and the R&D of a new neutron detectorbased on the Micromegas technology.

I - Summary description of the proposed TRADE project

In the European Roadmap towards the experimental demonstration of ADS (AcceleratorDriven System) [2], several experiments are indicated which should allow validating theseparated components of an ADS. This is the case for the accelerator (IPHI, TRASCO),the target (e.g. the MEGAPIE experiments), the sub-critical core (the FEAT, TARC andMUSE experiments [3]).Based on the original idea of Carlo Rubbia, presented at CEA in October 2000, a firstfeasibility report was produced on June 2001 by an ENEA (and partners) and CEAWorking Group and delivered to the ENEA/CEA management on July 2001.At the present time the finalisation of this feasibility report is scheduled for the nextsummer 2002.The project named TRADE (TRiga Accelerator Driven Experiment) is based on thecoupling of a 110 MeV, 2 mA proton cyclotron with a 1 MW reactor of type TRIGA,operating at the ENEA Casaccia site (Rome). The reactor core will be modified in orderto run the experiment at different levels of sub-criticality to explore the transition from an“external source” dominated regime to one dominated by core thermal-feedbacks.Moreover, an essential modification will consist of inserting at the center of the core atungsten target for the production of spallation neutrons.

II- Experiments

The experiments of relevance to ADS development to be carried out in TRIGA couldconcern:• The dynamic regime: the possibility to operate at some hundred kW of power and atdifferent sub-criticality levels (0.90 ÷ 0.99) will allow to validate experimentally thedynamic system behaviour vs. the neutron importance of the external source and to obtainimportant information on the optimal sub-criticality level both for a demonstrator and, byextrapolation, a transmuter.• Sub-criticality measurements at significant power.

• Correlation between reactor power and proton current. This correlation can be studiedat different sub-criticality and power levels.• Reactivity control by different means and possibly by neutron source importancevariation, keeping the proton current constant. In principle, this can be obtained changingthe neutron diffusion properties of the buffer medium around the spallation source (e.g.using different materials in the empty innermost fuel ring close to the target).• Start up and shut down procedures, including suitable techniques and instrumentation.The TRIGA experiments will benefit from the extensive experimental techniquesdevelopment performed in the zero-power MUSE experimental program. Most of theexperimental techniques will be directly applicable, in particular in the domain ofreactivity level measurement and monitoring. Even if the TRIGA experiments are madein a thermal neutron system, the frequency behaviour of the system transfer function issimilar to its behaviour in a fast system. This fact makes of the TRIGA experiment acrucial precursor of any future ADS larger size experiment.The possible different steps can be summarise in table 1

CONFIGURATION SOURCE KINETICS POWER

EFFECTS

TIME

MUSE DD/DT FAST NO à 2003TRIGA DD/DT THERMAL NO 2003à2006TRIGA SPALLATION THERMAL NO 2007TRIGA SPALLATION THERMAL YES 2007ADS SPALLATION FAST YES

As we can see from table 1, we approach the validation to the ADS by not changing morethan one parameter from one configuration to the next.

We propose to participate in the TRADE project and to perform the experimentalprogramme described in the two attached reports.

[1] The TRADE Collaboration. TRiga Accelerator Driven Experiment (TRADE)Feasibility Report. ENEA Report, March 2002.

[2] The European Technical Working Group on ADS, A European Roadmap forDeveloping Accelerator Driven System (ADS) for Nuclear Waste Incineration,April 2001.

[3] M. Salvatores et al., MUSE-1: a first experiment at MASURCA to validate thephysics of sub-critical multiplying systems relevant to ADS – 2nd ADTTConference, Kalmar, Sweden, June 1996.S. Andriamonje et al. Physics Letters B 348 (1995) 697 – 709 and J. Calero et al.Nuclear Instruments and Methods A 376 (1996) 89 – 103.H. Arnould et al. Physics Letters B 458 (1999) 167 – 180 and A. Abanades et alNuclear Instruments and Methods A 478 (2002) 577-730

NEUTRON SPECTRUM MEASUREMENT FOR TRADE PROJECT

S. Andriamonje, Y. GiomatarisCommissariat à l’Energie Atomique, CEA/DSM/DAPNIA, 91191 Gif-sur Yvette, France

Y. KadiEuropean Organization for Nuclear Research, CERN

CH-1211, Geneva 23, Switzerland

C. RubbiaEnte per le Nuove Tecnologie l’Energia e l’Ambiente, ENEA, 00196 Rome, Italy

I Introduction

In the TRADE (TRiga Accelerator Driven Experiment) project, several experiments areplanned for the validation of the safety and operational parameters. One of them is thedetermination of the neutron flux distribution in the core. Spectral information as afunction of distance from the source is needed to study the behaviour of different buffersand to be able to assess potential materials damage.Unfortunately the neutron spectra can not be obtained using only one type of detector. Tocover all the range of energy of the neutron produced, several detection techniques willbe used.

An example of several techniques used in TARC [1, 2, 3] to measure neutron fluencesover the desired energy range from thermal to MeV neutrons, is shown in figure 1. Thesetechniques can be used for the TRADE project.

Figure 1: Illustration of the energy ranges covered by the different detector techniquesused in TARC to measure neutron fluences.

6Li/233U detectors and a 3He detector in the scintillation mode covering the neutronenergy range from thermal up to about 100 keV and 3He ionisation detectors covering thehigher energy range from 10 keV to 2 MeV were used in TARC. These differentialmeasurements were complemented by measurements at specific neutron energies withtriple-foil activation methods. Several additional cross-checks of the neutron fluencewere performed, outside the energy range covered by the electronic detectors both at lowenergies (< 0.5 eV) using thermoluminescence techniques and at high energies(> 1.4 MeV) using fission measurement in 232Th. At higher energies threshold reactions[12C(n,2n)11C (En > 22 MeV) and 12C(n,3n)10C (En > 34 MeV)] were also used. Anexample of the result obtained from the TARC experiment is shown in figure 2.

For TRADE experiment, we propose in addition to use a conventional 3He detector.Three detection methods are proposed based on fission, (n,n’) reaction with Hydrogenand Helium and the activation foil.To have a small effect of the flux distribution the detector will be positioned in place of afuel rod.

Figure 2: Example of TARC measurements of the neutron fluence. E xdF/dE isshown as a function of neutron energy, from 2.5 and 3.57 GeV/c protons inhole number 10 (z = 7.5 cm, at a distance of 45.6 cm from the centre of thelead volume). The Monte Carlo predictions are shown as histograms. Thedata are from 6Li/233U detectors (open circles), 3He in the scintillationmode (full circles) and 3He in the ionisation mode (full squares). The errorbars include both statistical and systematic errors added in quadrature;

II Activation method

Activation foils are one of the most widely used types of neutron detectors because oftheir advantages: a) they take up comparatively little room; b) the foils may be activatedand the measurements made at one's convenience after the experiments; c) the presenceof gamma or other types of radiation does not interfere with the neutron activation of thefoils.Most materials are sensitive to three different energy regions: thermal region at lowenergies, resonance region at epithermal energies, and threshold region at epithermal andhigh energies. However, because the relative cross section of these processes can varyconsiderably from one material to another, some foils are better than others in certaincases. In our case, we are interested in epithermal neutron fluences so the resonancedetectors are more suitable.This technique is also known as the self-induction, foil sandwich, triple foil, or areaanalysis technique. A resonance detector has a very high (resonance) cross section atsome specific neutron energy and a relatively low cross section everywhere else. Theratio of these cross sections is often 1000 to 1 or more. Activating and countingcombinations of such foils may obtain the neutron fluence at defined energies (mainresonance energy).

II-1 Fast “Rabbit” technique

The technique is based on the gamma ray spectroscopy with GeHP detector. The neutronflux is deduced from the intensity of the typical gamma ray emitted by the radioactivenucleus formed after neutron capture. As well known this type of detector is verysensitive to neutron damage. The GeHP detector must be placed outside of the reactorenvironment. A pneumatic system named “fast rabbit” can be used. The sample is placedin an appropriate shuttle. The shuttle is then transferred by a pneumatic system from theirradiation port inside the core to a measurement port and inversely. An example of thistechnique used in the TARC experiment for the measurement of the transmutation rate of99Tc is shown in figure 3.

II-2 Possible measurements

Two “fast rabbit” system can be used:

a) The first one is dedicated to the measurement of the formed radioactive nucleushaving a short lifetime.Close to the spallation target where fast neutrons are produced, it is very interesting tomeasure the neutron flux as a function of energy (especially 16N and 17N activation inwater). For that the reactions 12C(n,2n)11C (threshold of 22 MeV) and 12C(n,3n)10C(threshold of 34.5 MeV) can be used. The lifetime of the formed nucleus 11C and 10Crespectively are too short (20 mn for 11C and 19 s for 10C), that leads to have a very shorttime for the transfer of the shuttle between the irradiation port and the measurement port.

An example of the gamma spectra obtained with the special carbon shuttle is shown infigure 4. The neutron flux are deduced from the number of counts under the 511 keV (the

result of the e+e– annihilation, following the + decay of 11C) and 718 keV (10C) lines.

Figure 3: a) Schematic layout of the ‘Fast Rabbit’ system; b) View of the carbon fibreshuttle containing the 99Tc sample in the irradiation port; c) View of thegamma-ray detection station.

Figure 4: Gamma-ray spectrum from the irradiated special 12C shuttle.

In addition at lower neutron energy, a natural silver sample (50% of 107Agà 108Ag(t1/2 = 2.37 minutes) and 50% of 109Agà 110Ag (t1/2 = 24.6 s)) and natural Aluminium

(27Al à 28Al (t1/2 = 2.2 mn)) , whose neutron capture properties are well known canalso be used.

b) The (n,2n) reaction rate of 232Th ( n,2n (En ) n (En)dEn , En ≥ 6 MeV) can

also be used for the determination of the neutron spectrum above 6 MeV. Themeasurement is based on the observation of the gamma rays emitted by 231Th formed bythe (n,2n) reaction on 232Th. As is well known, natural 232Th is an alpha emitter and a lotof gamma rays are emitted during this de-excitation. The gamma rays from 231Th aresuperimposed on the natural gamma ray background of 232Th. The intensity of thegamma rays from 231Th is extracted from the subtraction of this background. Thisintroduces an uncertainty for the determination of the n,2n reaction rate.For TARC experiment, a special experimental set up was arranged to observe the25.6 keV gamma ray transition of 231Th. This line has the advantage of a relatively highintensity (14.5%), and lies in the gamma ray region where the contamination is unlikely(figure 5).

Figure 5: Observed photons from the irradiated and non-irradiated 232Th sample inthe energy range from 21 keV to 40 keV.

The following three conditions are needed for the observation of this γ-line:– the sample must be thin enough to minimise the gamma autoabsorption and the selfshielding of the incident neutrons– the amount of 232Th must be sufficient to allow an irradiation in the neutron flux in areasonable time (about 12 h)– a high-resolution and high-efficiency X-ray detector must be used.

Since 231Th is a beta emitter with 25.52 h of half-life the use of the fast rabbit system isnot necessary. A second pneumatic system but very simple compared to the first one canbe used. It is also very useful to have this second pneumatic for handling the sample usedfor other activation foil measurements. For example: - the measurement of theproduction rate of 233U from 232Th and of 239Pu from 238U with the use of delayed γspectroscopy, - triple foil activation method for In, Au, and W elements.

III - Neutron detector based on Micromegas concept

A novel neutron detector based on the Micromegas concept, has been developed byDAPNIA group. The origin of this development is to have a profiler detector for n_TOFfacility at CERN [4]. This type of detector can be used for neutron detection: theconventional neutron reaction of 6Li (10B), fission fragment detection and the detection ofH+ and He+ from the (n,n’) reaction of Hydrogen and Helium. A detailed description ofthe Micromegas detector can be found in references [5 ] and[ 6].It is a double-stage parallel plate chamber, consisting of a conversion gap and anamplification gap, separated by a micromesh. Ionisation electrons, created by the energydeposition of an incident charged particle in the conversion gap, drift and can betransferred through the cathode micromesh; they are amplified in the small gap, betweenanode and cathode, under the action of the electric field, which is high in this region. Theelectron cloud is finally collected by the anode microstrips, while the positive ions aredrifting in the opposite direction and are collected on the micromesh.

In order to operate the MICROMEGAS detector as a neutron detector for the n-TOFfacility at CERN, an appropriate neutron/charged particle converter must be employed(see figure 6) which can be either the detector's filling gas or target with appropriatedeposit on its entrance window [7].

Figure 6: The principle of the Micromegas concept for neutrons detector ([8])

Since the neutron energy range of the n-TOF facility extends from 1 eV to 250 MeV,there is not a unique choice of an efficient converter. Inter-dependent parameters such asthe high neutron reaction cross section, the low charged particles energy loss inside theconverter, their subsequent energy-angular distribution and the range inside the filling gashas been considered and optimised.

For nTOF detector, the neutron/charged particle converter employed are:6Li(n,α) for neutron energy up to 1 MeV [4] and H(n,n’)H and 4He(n,n’)4He for highenergy neutron) which are the detector's filling gas. Three type of gas mixtures have beenused: Ar + iC4H10 (2%), He + CH4 (10%) and He + iC4H10 (3.8%). The percentage of

iC4H10 and CH4 has been chosen to be low in order to respect the non-flammabilty of the

mixed gas [8].A similar detector can be constructed for TRADE experiment. This detector must besmaller so that it can be put inside the specified rod of TRIGA reactor.Fissionable elements such as 232Th, 235U, 238U, 237Np and 239Pu can be used aneutron/charged particle converter in place of the 6Li .

The energy dependence of the reaction rate is obtained as a function of the differentfission threshold of the considered element.The element mentioned above are α emitters, the amplitude of the Micromegas pulseprovided from the α and from the fission fragments are very different, the determinationof the fission rate is easy.A more detailed description of the proposed detector is developed in the next attachedreport.

IV - Solar cells detectors

An additional type of detector can also be used for counting the fission fragments. Thatconcerns the use of solar-cell semiconductor detectors and detected fission fragmentsfrom a thin (≈ 1 mg/cm2) fissionable sample.

Inexpensive slabs of polycrystalline silicon normally used as solar cells can be cutinto chips of ≈ 1 cm2 and can function as a diode giving a good signal for fissionfragments. An examples used in FEAT [9] experiment are shown in figure 7. Thesewere encased in boxes together with the Uranium converter deposits and mounted on 10rods, each holding 16 detectors as shown in figure 7. They could be displaced verticallyin-situ, relative to the converter foils, during the data taking in order to detect thebackground level coming from other interactions than fission which was found to benegligible. Another arrangement also shown in figure 7 was that of detectors locatedinside the Uranium rods.

Figure 7: Photovoltaic silicon solar cells were used to detect fission fragments from

thin Uranium converter foils. Measurements were done in water by rods

comprising 16 detectors each (a). Measurements were also done inside

Uranium bars by inserting detectors between U cartridges (b).

V Conclusion

Using the different methods described in this proposal, the neutron spectrum fromthermal up to several MeV relative to the TRADE project can be obtained. The estimatedpower is 1 MWatt, the experimental condition is different compared with TARC andFEAT, in particular concerning the temperature of the detector environment (up to 300degree C). A special study and test of the material which will be used for the detectorconstituent (cell, cabling etc) must be performed.

References:

[1] H. Arnould et al, “Experimental Verification of Neutron Phenomenology in Leadand Transmutation by Adiabatic Resonance Crossing in Accelerator DrivenSystem” Phys. Lett. B 458 (1999) 167

[2] A. Abanades et al, “Experimental verification of neutron phenomenology in leadand transmutation by adiabatic resonance crossing in accelerator driven systemsA summary of the TARC project at CERN”, Nucl. Instr. & Meth. A 463 (2001)586

[3] A. Abanades et al, “Neutron-driven nuclear transmutation by adiabatic resonancecrossing TARC”, Project report “Nuclear Science and Technology” EuropeanCommission EUR19117 EN (1999)

[4] C. Rubbia et al., “A High Resolution Spallation Driven Facility at the CERN--PSto Measure Neutron Cross Sections in the Interval from 1 eV to 250 MeV”,CERN/LHC/98-02 (EET) (1998) and “a Relative Performance Assessment”,CERN/LHC/98-02 (EET)-Add. 1, Geneva, 15 June 1998.

[5] Y. Giomataris et al., “A high-granularity position sensitive gaseous detector forhigh particle-flux environments”, Nucl. Instrum. Methods A376 (1996) 29

[6] G. Charpak et al., “First beam test results with micromegas, a high rate, highresolution detector”, CERN LHC/97-08(EET), DAPNIA-97-05

[7] S. Andriamonje et al, “Experimental studies of a MICROMEGAS neutrondetector”, Nucl. Instrum. Methods A 481 (2002) 36-45

[8] S. Andriamonje et al, “The Micromegas neutron detector for CERN n_TOF”to be published, Nucl. Phys. B, Proceedings of the International Conferenceon Advanced Technology and Particle Physics, Como-Italia (2001).

[9] S. Andriamonje et al , “Experimental determination of the energy generated innuclear cascades by a high energy beam”, Phys. Lett. B 348 (1995) 697

Micromegas detector for in-core measurements

S. Andriamonje, Y. Giomataris, J. Pancin,Commissariat à l’Energie Atomique, CEA/DSM/DAPNIA, 91191 Gif-sur Yvette, France

G. Bignan, G. Imel, C. JammesCommissariat à l’Energie Atomique, CEA/DEN/DRNCAD, CEA/Cadarache

13108 Saint-Paul-Lez-Durance Cedex France

P. Cennini, Y. KadiEuropean Organization for Nuclear Research, CERN

CH-1211, Geneva 23, Switzerland

I - Introduction

From a recent results [1.2], it is demonstrated that a Micromegas detector is excellent forneutronic studies at a very large energy range from thermal up to 250 MeV. The detectorcan be used not only for a neutron beam profiler but also for neutron flux measurementfor research with neutrons. An example is shown for the determination of thecharacteristic and performance of the CERN n_TOF facility [2]. For this experiment, anappropriate neutron/charged particle converter:(i) 6Li(n,α) for the neutron having an energy up to 1 MeV, (ii) H(n,n')H and 4He(n,n')4Hefor higher energy neutron, have been used (see figure 1).This method can be extended for the measurement of neutron flux in-core environment.A fissionable element such as 235U, 238U, 232Th or other can be used for neutron/chargedparticle converter.

The aim of this propose is the new development of detector used in particle physics forthe use in reactor environment where the condition is very different, the presence of veryhigh background of neutron and gamma and the operation in high temperature. One ofthe main qualities needed for that detector is his high resistance to the radiation. Severalexperiments show that Micromegas detector meet with this requirement.In opposite of the usual Micromegas detector used in particle physics experiment wepropose to develop a compact and very small detector (2cmx2cmx2cm).

II - Description of MICROMEGAS

The general description of the principle of Micromegas (MICRO-MEsh-GAseousStructure) technology can be found in references [3] and [4]. We describe here the typicalaspect of the proposed Micromegas detector for in-core measurement in reactor.An example of Micromegas neutron detector used in n_TOF experiment [2] is reported infigure 1. As shown in this figure, the amplification occurs between the mesh plane andthe microstrip plane. A small gap, of about 100 µs, between the anode and cathode plane

is kept by precise insulating spacers. The device operates as a two-stage parallel plateavalanche chamber.A micromesh separates the conversion space, of about 3 mm, from a small amplificationgap that can be as small as 100 µm. This configuration allows to obtain, by applyingreasonable voltages in the three electrodes, a very high electric field in the amplificationregion (about 100 kV/cm) and a quite low electric field in the drift region.Therefore, the ratio between the electric field in the amplification gap and that in theconversion gap can be tuned to large values, as is required for an optimal functioning ofthe device. Such a high ratio is also required in order to catch the ions in the smallamplification gap: under the action of the high electric field, the ion cloud is quicklycollected on the micromesh and only a small part of it, inversely proportional to theelectric field ratio, escapes to the conversion region.

Figure 1: The principle of Micromegas for neutron detector used in n_TOFexperiment [2]

Ionization electrons, created by the energy deposition of an incident charged particle inthe conversion gap, drift and can be transferred through the cathode micromesh; they areamplified in the small gap, between anode and cathode, under the action of the electricfield, which is high in this region. The electron cloud is finally collected by the anodemicrostrips, while the positive ions are drifting in the opposite direction and are collectedon the micromesh. The electric field must be uniform in both conversion andamplification spaces. This is easily obtained by using the micromesh as the middleelectrode.Figure 2 shows a schematic representation of the proposed detector. It consistsof the following components:

- The anode electrode is made of Copper pads of 150 µm, with 5 mm x 5 mm pads of 1mm gap, are printed on a thin substrate. Each pad (S1, S2, S3, S4) is read-outindividually by a low noise preamplifier.

- The micromesh is a metallic grid, 3 micron thick, with 37 µm openings every 50 µm. Itis made of nickel, using the electroforming technique.

- The conversion-drift electric field was defined by applying negative voltages on themicromesh (HV2) and a slightly higher voltage on a second electrode (HV1), spaced by 1mm in order to define a conversion--drift space. This second electrode made by a 20 µmAluminium foil, will be equipped with four neutron/charged particle converter which are:6Li (or 10B), Empty, 235U and fissile element with fission threshold (see figure 2).

- The various elements of the parallel-plate structure were placed in a stainless steel boxwith an 3 cm x 3 cm x 3 cm to respect the internal diameter (3.6 cm) of the TRIGA rods.The box will be filled by a gas mixture of He + (3.8%)Iso-butane at atmosphericpressure. The proportion of Iso-butane is chosen to provide a non-flammable gas mixture.

Figure 2: Schematic view of the micro-Micromegas detector for neutron fluxmeasurement.

III - Construction and test of the detector

In TRADE project the estimated power is 1 MWatt, the experimental condition isdifferent compared with particle and nuclear physics experiments. We are in presence ofthe very high neutron flux and high background of gamma rays. The temperature in thecore is estimated to be about 3000 degree C. Therefore a special study and test of thematerials used for the various element of the detector (frames, cables, etc…) must beperformed. The experiences of the DEN Cadarache physicist will be valuable.

Usually a flowing gas is used in Micromegas, but in the present case it is not convenientto have a long gas tube for this task. A test of the detector placed in sealed box filled byappropriate mixed gas and in operational pressure is needed. The possibility to use aflowing gas with a very small and special bottle of the mixed gas will be envisaged. Thismethod is already used in space experiment. That permits a compensation of the possiblesmall leak in the detector box.The construction of the drift electrode equipped by 6Li, 235U and other fissile elementswill be performed with DEN Cadarache collaboration. The experience acquired for thedevelopment of Cadarache ionisation chamber [5] can be used directly for the design andconstruction of the proposed special Micromegas detector.The combination of both type of detector such as a ionisation fission chamber actuallyused in MUSE [6] experiment and the new proposed Micromegas permit to obtain ingood conditions the neutronic parameters needed in TRADE project. A large part of theneutron spectra can be obtained. An example of the first preliminary results from then_TOF experiment are shown in figure 3. In this figure are reported the time converted toequivalent neutron energy of the relative flux seen by Micromegas detector. Two plotsare reported in figure 3, the result obtained from the 4He + (3.8%)iC4H10 and Ar +(2%)iC4H10 respectively.

Figure 3: Relative flux (counts rate) of the n_TOF neutron beam seen byMicromegas detector as a function of the neutron energy.

Figure 3 shows clearly the 1/v response at lower energies and the resonance at 250 keVcharacteristic of the 6Li converter. At higher energy (several MeV) there is a large bumpas a consequence of the increase of neutron fluence in this energy (spallation peak) andalso the increase of detector efficiency provided by the detection of recoil nuclei from thefilling gas. Several deeps are observed in the energies corresponding to thedifferent elastic resonances of the 16O. This is the result of the dispersion of neutron ofthose energies from the main beam by the oxygen present in the water used for cooling ofthe spallation lead target. The shift, observed in the figure between the two plots, is theresult of the presence of the elastic resonance of the 4He(n,n')4He reaction.

The use of the mixed gas of 4He + iC4H10 combined with the four neutron/chargedparticle mentioned before allows to obtain a large part of the neutron spectra of theTRADE project.One of the advantages of Micromegas is the possibility to measure simultaneouslyneutron, alpha and fission. In addition, at low neutron flux it is possible to count one byone the incident particle using a low noise preamplifier. At very high counting rate (~ 1GHz) only a continuous current measurement can be used.

IV - Conclusion

At the present time, the all Micromegas detector used in physics experiment have a largedimension. The development of the micro Micromegas for neutron detector, able todetect a broad neutron energy is not only the normal way for detector R&D but also forthe future applications (reactor flux, new project like TRADE , etc…).The co-operation between reactor physicists and particle physicists will be very useful forthis type of detector development.

References

[1] S. Andriamonje et al, “Experimental studies of a MICROMEGAS neutrondetector”, Nucl. Instrum. Methods A 481 (2002) 36-45

[2] S. Andriamonje et al, “The Micromegas neutron detector for CERN n_TOF”to be published, Nucl. Phys. B, Proceedings of the International Conference onAdvanced Technology and Particle Physics, Como-Italia (2001).

[3] Y. Giomataris et al., “A high-granularity position sensitive gaseous detector forhigh particle-flux environments”, Nucl. Instrum. Methods A376 (1996) 29

[4] G. Charpak et al., “First beam test results with micromegas, a high rate, highresolution detector”, CERN LHC/97-08(EET), DAPNIA-97-05

[5] G. Bignan et al Bignan G. and Guyard J.C. (1994), Sub-miniature FissionChamber with Tight Feedthrough, Patent CEA N°94-14293.Poujade O. and Lebrun A. (1999), Modeling of the Saturation Current of aFission Chamber taking into Account the Distortion of Electric Field due to SpaceCharge Effects, Nuclear Instruments and Methods in Physics Research A, 443,pp. 673-682.

[6] M. Salvatores et al., MUSE-1: a first experiment at MASURCA to validate thephysics of sub-critical multiplying systems relevant to ADS – 2nd ADTTConference, Kalmar, Sweden, June 1996.



Title: COMPASS

Date of the first CSTS presentation: 1997

Experiment carried out at: CERN SPS

Spokes person(s): F. Bradamante (University Trieste) and S. Paul(T.U.München)

Contact person at SPhN: A. Magnon

Experimental team at SPhN: J. Ball, Y. Bedfer, C. Bernet, F. Kunne, J-M. Le Goff,

J. Marroncle, C. Marchand, D. Neyret


SPhN (8), SEDI (10 man x year), SIS (3 man x year)


Bielefeld, Bochum, Bonn (ISKP and PI), Burdwan and Calcutta, CERN, Dubna (LPP and LNP),

Erlangen, Freiburg, Heidelberg, Helsinki, Mainz, Moscow (INR, LPI and State University), München

(LMU and TUM), Nagoya, Protvino, Saclay, Tel Aviv, Torino (University and INFN),

Trieste (University and INFN), Warsaw (SINS and TU), 208 physicists in all

SCHEDULE

Starting date of the experiment [including preparation]: 1997

Total beam time allocated: 2.5 years

Total beam time used: 100 days

Data analysis duration: undetermined

Final results foreseen for: First results in 2002

BUDGET Total Already Used

Total investment costs for the collaboration: 30 MCHF (*)

Share of the total investment costs for SPhN: 11 MFF (+) 9,2 MFFTotal travel budget for SPhN:

(*) Memorandum on current status of funding of COMPASS - NA58 Collaboration. CERN-SPSLC-96-55 ; SPSLC-M-592. - Geneva : CERN , 1997.(+) Investment and running costs.


1 COMPASS

COMPASS [1] is a high-energy physics experiment at the Super Proton Synchrotron (SPS)at CERN aimed at the study of nucleon structure on the one hand and on the other handhadron structure and spectroscopy. These two distinct physics programs require dierentbeam and target congurations, but make use of a common spectrometer apparatus.

The experiment was approved by CERN authorities in 1997. Its experimental apparatushas been constructed by an international collaboration, with important and innovativecontributions from DAPNIA. It is ready to enter production data taking in 2002, with areduced setup, the rst production runs being dedicated to the nucleon structure aspectof the experiment. The main objective in this respect, the determination of the gluonpolarisation in the nucleon, G, is delineated infra. An overview of the spectrometerand of the main contributions of DAPNIA to its apparatus is then given. Finally a briefaccount of the 2001 run and of the prospects for 2002 is presented.

2 G Measurement

The main objective of COMPASS in nucleon structure physics is the measurement of G.This measurement has emerged as the most promising answer to the puzzle posed by aseries of recent experiments carried out at CERN, SLAC and DESY, which concluded tothe smallness of the contribution of the quark spins to the nucleon spin.

COMPASS intends to access G by the study of the semi-inclusive cross-section foropen charm production. This channel is selected because the production of charm quarkhadrons is triggered in leading order by the photon-gluon fusion mechanism g ! qq,shown in Fig.1, and is therefore sensitive to the gluon density inside the nucleon, whereasthe negligible charm content of the nucleon at available Q2 values excludes competing nongluonic processes.

p

µ

c

cq

γ *

µ

k

G

Figure 1: The cc production through Photon Gluon Fusion in p scattering

The determination of G=G is obtained by the measurement of the cross-section asym-metry in the scattering of polarised muons o a polarised target. The tagging of charm

events is based on the identication of D0 mesons via their D ! K decay channel.

The sensitivity of COMPASS to G=G by this method is expected to be Æ(G=G) ' :11.The experiment will also explore other ways of accessing this quantity, e.g. via theproduction of oppositely charged hadron pairs at high PT , as well as others nucleonstructure observables, e.g. the transversity structure function, h1

3 Spectrometer

The measurements outlined supra require a high energy high intensity polarised muonbeam (provided by an upgraded SPS M2 beam line able to deliver 2:108 160 GeV perSPS spill), a polarised nucleon target (the solenoid, built by DAPNIA, and target, lledwith LiD, of the former SMC experiment at CERN SPS, are used) and a spectrometer.

The spectrometer is two-staged, for large angle low momentum and small angle largemomentum particle reconstruction. Both stages conform to a same pattern, with a dipolemagnet, trackers and particle identication. The large angle stage is shown on Fig.2, inits original design, together with the polarised target solenoid. Its particle identicationsection is complete, with a RICH detector, electromagnetic and hadronic calorimetersand a muon lter. The tracking detection in the upstream section is the responsibility ofDAPNIA, except for the coverage of the beam area.

Figure 2: The Large Angle Stage of the spectrometer. The Micromegas detectors and drift

chambers built by DAPNIA serve as trackers on either side of the SM1 dipole magnet.

In its inception, the project was to take charge of the zone between the target and themagnet. In this zone the conditions in term of particles rate are extremely severe, for thefull spectrum of particles produced in the target is present, in so far as the low energycomponent is not yet swept away by the magnetic eld, and depends strongly on thedistance to the beam. To cope with these constraints we have devised a tracking systemdivided into a set of two nested detectors with dierent rate capabilities: for the outer-most region drift chamber detectors are used, while for the innermost region Micromegashigh rate microstrip detectors were retained.

In the present setup, Saclay drift chambers are also used to supplement the large areatracking behind the magnet.

4 Micromegas

Large area Micromegas detectors of 4040 cm2 have been developed to meet the specialneeds of the experiment [2][3][4][5], together with dedicated ASIC (SFE16[6]) front-endelectronics. In addition to the above mentioned rate requirements the detectors had to ful-ll requirements set on their position accuracy and on their total mass, which determinesthe magnitude of the multiple scattering. The performances achieved, under particlesrates of 20 MHz integrated over the detector area and up to 90 kHz per channel, arethe following:

Spatial resolution: 70 m.

Time resolution: 9 ns.

Mass: 0:2 % of radiation length per plane.

12 detectors have been built.

Figure 3: A doublet of Micromegas detectors.

5 Drift Chambers

Away from the beam, beyond the 4040 cm2 covered by the Micromegas, the particlesrate is still very challenging for the drift chamber technique. To overcome this constraint,

we have settled on a 7 mm wide drift cell, which results in a maximum events rate of300 kHz per channel. A fast Ar=C2H4=CF4 gas mixture was selected and dedicatedfront-end PC boards have been developed. The resolution achieved is 150 m. In all, 3detectors of 120120 cm2, with 8 coordinates each, have been built.[7]

Figure 4: Drift Chamber.

6 6LiD Polarized Target

The 6LiD material has been chosen for its high dilution factor (50%). In 2001 it has beenpolarized at a temperature of 250 mK, in the eld of the solenoid previously build bythe DAPNIA for the SMC experiment. It is the rst time that such a volume of 6LiD (1liter) could be polarized to a high value (50%) at the eld of 2.5 T . This success is dueto : 1/ Progress in the fabrication and irradiation processes at the Bochum institute; 2/High homogeneity (3.105) of the SMC solenoid eld.

The target has been continuously operated during 3 months in 2001, with polarizationreversals 3 times a day.

7 Run 2001 and Prospects for 2002

The 3 months running period of 2001 was put to prot to commission most of the de-tectors, and other building blocks of the experiment. For what concerns DAPNIA, 6Micromegas, out of 12, and one drift chamber, out of the 2 originally planned, were

installed. For both commissioning was successful: all encountered problems, either in-fantile diseases in the production of Micromegas PC boards or higher noise level amongthe 6000 electronics channels installed in all, have been progressively mastered. All14 detectors are scheduled to be ready and installed for the beginning of 2002 data tak-ing. An third drift chamber, is being build and is foreseen to be delivered for mid run.For what concerns COMPASS in general, the prospects for 2002 are promising: updatedevaluations for the precision on G=G [8] conrm the values cited in the proposal. Onehas to contend though that the reductions of beam time (by 25 %) decided by CERNauthorities to make for the LHC funding crisis, will aect COMPASS in the long term,and so more so as it is a statistics limited experiment.

References

[1] COMPASS Proposal, CERN/SPSLC/P297, March 1, 1996.

[2] D. Thers et al., NIM A 469 (2001) 133-146.

[3] D. Thers, PhD Thesis, Clermont-Ferrand (2000), DAPNIA/SPhN-00-06-T 02/2001.

[4] Ph. Abbon et al., NIM A461 (2001) 29.

[5] A. Magnon et al., NIM A478 (2002) 210-214.

[6] E. Delagnes et al., IEEE Trans. Nucl. Sci. NS-47 (2000) 1447.

[7] H. Pereira, PhD Thesis, Paris XI Orsay (2001), DAPNIA/SPhN-01-02-T 11/2001.

[8] J.-M. Le Go, http://wwwcompass.cern.ch/compass/notes, COMPASS note 2002-02.

Documents

Conseil Scientifique et Technique du Service de Physique